Serology, chemistry, and genetics of O and K antigens of Escherichia coli.

BACTERIOLOGICAL Rzvizws, Sept. 1977, p. 667-710 Copyright 0 1977 American Society for Microbiology

Vol. 41, No. 3 Printed in U.S.A.

Serology, Chemistry, and Genetics of 0 and K Antigens of Escherichia coli IDA 0RSKOV, FRITS ORSKOV,* BARBARA JANN, AND KLAUS JANN Collaborative Centre for Reference and Research on Escherichia (WHO), Statens Seruminstitut, 2300' Copenhagen S, Denmark,* and Max-Planck-Institut fur Immunbiologie, Freiburg, West Germany

INTRODUCTION.............................................................. 668 HISTORY ..........................................---------668 SEROLOGY OF 0 ANTIGENS ................................................. 668 Techniques for O-Antigen Determination ...................................... 668 Test 0 Antigens .....................................-.------------...-.669 Cross-Reactions Between 0 Antigens of E. coli ............................. 669 Cross-Reactions Between 0 Antigens of E. coli and Shigella, and Salmonella and KiebsieUa .....................................................-.-.669 Cross-Reactions Between E. coli 0 Antigens and Antigens Outside the Enterobacteriaceae Group .669 SEROLOGY OF K ANTIGENS .674 Classical L A B Subdivision in the Light of Recent Findings .674 Polysaccharide K Antigens .677 Techniques for K-antigen determination .677 Test K antigens............................................................ 678 Cross-reactions between E. coli K antigens .679 Cross-reactions between E. coli 0 antigens and K antigens .679 Cross-reactions between E. coli K antigens and polysaccharide K antigens outside the Escherichia group .679 M Antigen .681 Protein K Antigens (Fimbriae Classified as K Antigens) .681 Fimbriae Not Classified as K Antigens .681 CHEMISTRY OF 0 ANTIGENS .682 Isolation and Purification .682 Chemical Characterization .682 Neutral Polysaccharide Chains .683 Acidic Polysaccharide Chains .683 CHEMISTRY OF R ANTIGENS .683 R Antigens with a Complete Core .684 R Antigens with an Incomplete Core .685 GENETICS OF 0 AND R ANTIGENS .685 CHEMISTRY OF K ANTIGENS .688 Isolation and Purification of Polysaccharide K Antigens .688 K Antigens Occurring in 0 Groups 08, 09, and 0101 .688 K Antigens Occurring in 0 Groups Other than 08, 09, 020, and 0101 .689 M Antigen .690 GENETICS OF K ANTIGENS .691 K Antigens Determined near his .691 K Antigens Determined near serA .692 M Antigen ...................................................................-692 K Antigens Determined by Plasmids ..................................-. 692 BIOSYNTHESIS OF 0 AND K ANTIGENS .692 MULTIPLE OCCURRENCE OF POLYSACCHARIDE ANTIGENS .693 PATTERNS OF POLYSACCHARIDE ANTIGENS .693 CHEMICAL BASIS OF ANTIGENIC SPECIFICITY ............................ 694 CHEMICAL BASIS OF CROSS-REACTIONS .695 PREVALENCE OF SEROLOGICAL GROUPS AND TYPES OF E. COLI IN PATHOLOGICAL CONDITIONS IN HUMANS AND ANIMALS .695 E. colU from Humans .695 E. coli from Cattle, Pigs, and Poultry ...................................... 698 SUMMARY AND CONCLUSION .698 LITERATURE CITED .699 667

668

ORSKOV ET AL.

BACTERIOL. REV.

gen found in all smooth (S forms) EnterobacteINTRODUCTION It is thermostable in the sense that the Several reviews and monographs that cover riaceae. keep the immunogenic, agglutinating, bacteria some of the themes of this review have been capacity after boiling. and agglutinin-binding 192, published in the last decade (39, 112, 155, 0 polysaccharide is the The antigen O-specific 279). An increased interest in Escherichia coli (LPS). Muin recent years from both human and veteri- of the cell wall lipopolysaccharide 0 specificity are R have lost the tants that nary medicine has been followed by an interest is R of little or no value forms. As serology in the surface F&>ructures of these bacteria be- serotyping, it will not be discussed here. for cause of their special role in pathophysiological proceszs. their usefulness in epidemiological Techniques for O-Antigen Determination studies, and their importance for the normal host. We therefore immunological status of the Sera for 0 determination are produced by agree that the present review covering all ofthe immunization of rabbits with cultures heated immunogenic surface structures, with empha- at 1000C for 2 h. Broth cultures or agar plate sis on the K antigens, is needed. It is based on suspensions heated at 1000C for 1 h are used as new methods and insight and should thus hopeantigens. With these two ingredients, one will fully give a coherent description, taking into have an agglutination system that, to a large consideration structure, chemistry, genetics, extent, involves only one antigen and its hoand serology ofthe relevant surface antigens. mologous antibody, and therefore the bacterial agglutination method is very simple and sensiHISTORY tive for qualitative 0 determinations. Some In this section only few fundamental data mucoid E. coli strains with a special heat-rewill be presented. The first successful attempt sistant capsular antigen will only be agglutinato classify E. coli by serological methods was ble in 0 serum after autoclaving (1200C for 2 h). carried out by Kauffmann (127), who was able In the manual on Enterobacteriaceae by Edto subdivide E. coli into a number of 0 groups. wards and Ewing (39), detailed descriptions of Using a boiled culture for 0-antiserum produc- procedures for 0-antiserum production and 0tion and as an antigen in agglutination tests, he antigen determination can be found. Orskov was able to establish well-defined Escherichia and Orskov (191) recently published another 0 groups. The first antigenic scheme, comprisdetailed account of general O-antigen determiing 20 0 groups, was thus established by Kauff- nation. The test 0 sera, at present, are colmann (128). Knipschildt (138) added another 5 lected into pools, each containing cross-reacting 0 antigens, the scheme contained 25 0 0 sera. An unknown strain is first examined in groups. Since then, many 0 antigens have been the pools and, if positive in one pool, it is examadded, and the antigenic scheme presented ined in the single sera of that pool. Finally, it is here comprises 164 0 groups (see below). titrated in that serum (those sera) that has During his investigations Kauffmann ob- given a positive reaction. In most laboratories served that many freshly isolated strains were 0 determination of E. coli strains is carried out not agglutinated when examined in the non- in tubes or trays (microtiter system), but may heated state in 0 antiserum. A similar phe- of course also be carried out on slides, if only nomenon was well known, e.g., from Salmothe antigen has been heat treated at 1000C nella typhi, which is also nonagglutinable in 0 before use. If many strains have to be examserum (O inagglutinable) when equipped with ined, it may be advantageous to use a more or the Vi surface antigen; this inhibition of agglu- less automatized agglutination method. Such tinability in 0 serum could be overcome by heat methods are described in some recent publicatreatment. It was similarly possible to describe tions (8, 69, 82). a series of additional surface E. coli antigens Bacterial agglutination is still the best and with somewhat different physical and serologi- most simple technique for O-group determinacal characters, which all caused inagglutinabil- tion of E. coli, although other techniques are ity in 0 serum that could be overcome by heat also used. Indirect (passive) hemagglutination, treatment. For a more detailed discussion of which inadvertently led to the discovery of the these antigens, the L, A, and B antigens, later enterobacterial common antigen, was used by collectively called the K antigens, see below. Kunin and Beard (144) and Kunin et al. (145). For a detailed first-hand description of the his- Different immunoprecipitation tests may also torical development of E. coli serology, see be employed for O-grouping purposes, but, Kauffmann (130). probably because of the simplicity and reliabilof the bacterial agglutination test, gel preity SEROLOGY OF 0 ANTIGENS cipitation has not been used routinely for 0The 0 antigen is a thermostable surface anti- group determination. However, gel precipitaso

VOL. 41, 1977

tion in two dimensions and especially immunoelectrophoresis (219, 241) have been used extensively in many 0 analyses and comparative examinations (19, 101, 102, 194, 203, 204). Immunoelectrophoresis will also give information on the electric mobility of the O-antigen molecule (194). Test 0 Antigens Table 1 gives an antigenic scheme, listing test strains for all established 0, K, and H antigens arranged according to their 0 antigens, with the origin of the strains and the relevant references included. It should be stressed that this is a restricted antigenic scheme, i.e., that the 0, K, and H antigens presented have been found in many other combinations and, furthermore, that antigens listed with the same symbol may be and in many cases are nonidentical. It is also important to remember that new strains are often found which, even though they can be assigned to a certain existing 0 group, will show crossreactions not described hitherto. The first 0 groups, 01 to 0110, were established in the forties by Kauffmann and Knipschildt (see Kauffmann [130, 132]). However, because K and H antigens were not examined in most strains, 026 to 0110 were never published in a formal antigenic scheme. Some 0 groups are cancelled: 031 and 047 because of near identity to other groups and 067, 072, and 094 because they turned out to be Citrobacter freundii. Among the first 110 0 groups are many that are only found infrequently, the reason being that practically any E. coli strain that could not be grouped with available 0 sera at that time was established as a new 0 group. At that time -in the beginning of the forties there was only scanty interest in medical appli.; cation of E. coli typing. Later, when it was found that some well-defined serotypes were associated with outbreaks of infantile diarrhea, this attitude was changed, and serotyping has since been applied to many different problems within medical microbiology. It has been necessary, however, to restrict the number of newly established 0 antigens in order to keep the number of test antigens within a handy size. Therefore, only strains that have a special importance from a medical, epidemiological, or scientific point of view have been established as new 0 test antigens. It has also been considered important that any new test antigen, before being officially established, is examined in at least two laboratories equipped with all test antisera and test strains. Among the strains with numbers higher than 110, several will be found that have been isolated from pathological conditions in animals.

0 AND K ANTIGENS OF E. COLI

669

Cross-Reaction Between 0 Antigens of E. coli Many cross-reactions exist between the single 0 antigens, and a great number of crossabsorbed 0 sera (factor sera) are necessary for a precise 0 diagnosis. However, even using these absorbed sera, we are unable to tell whether an unknown strain, assigned to a certain 0 group, is really O-antigen identical with the test strain of that 0 group. Only serum production with the unknown strain followed by mutual cross-absorption can give the definite answer. Table 2 shows the stronger cross-reactions between all established O-antigen test strains and the corresponding specific 0 antisera. These results are based primarily on the results of Orskov and Orskov in Copenhagen, but they also include data from the examinations of Ewing (39, 50). Cross-Reactions Between 0 Antigens of E. coli and Shigella, and Salmonella and Klebsiella Many O-antigenic cross-reactions have been described between E. coli and Shigella 0 antigens. Ewing et al. (46) have gathered data that show that only two Shigella serovars are not related to one or more of the E. coli antigens 01 to 0148. He further shows that a number of sub judice Shigella serovars are related to E. coli 0 groups. 0 relationships betweenE. coli 0149 to 0163 and Shigella have been examined by Rowe et al. (237). These findings stress the well-known close relationship between the two genera. A number of invasive E. coli strains described in recent years, which can give rise to dysentery-like disease, have 0 antigens closely related to Shigella 0 antigens (Table 3). Kauffmann (130, 132) and Frantzen (57) have described O-antigen cross-reactions between E. coli and Salmonella, and recently Refai and Rohde (?29) have carried out similar investigations with 142 E. coli O-antigenic test strains in available Salmonella and Arizona sera. Table 4 shows the results of Refai and Rohde together with the earlier results in a condensed form. Cross-reactions between 0 antigens of E. coli and Klebsiella pneumoniae have been described by Kauffmann (129) and 0rskov (198) (Table 5). Most of these cross-reactions are mutual, but identity is not found in all cases. Cross-Reactions Between E. coli 0 Antigens and Antigens Outside the Enterobacteriaceae Group Winkle et al. (307) described 0-antigen relationships between Vibrio cholerae and E. coli, Salmonella, and Citrobacter strains. Springer (269, 271) has analyzed the cross-reactions between E. coli 086 antigen and human blood

670

ORSKOV ET AL.

BACTERIOL. REv.

TABLE 1. E. coli antigenic scheme comprising all 0-, K-, and H-antigenic test strains February 1977; arranged according to O-antigen numbers 0

lb 1 2 2 2 2 2 3 3 3 4 4 4 4 5 6 6 6 6 6 6 7 7 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8

8 8 8 8 8, 60 9 9a 9ab 9 9 9 9 9 9 9 9 9 9

K

H

Culture no.

1 51 1 7 (56)C 1 2 NEd 2ab NE NE 3 6 12e 52 4 2ac 13 15 53 54 13 1 7 8 25 27(A) 40 41 42(A) 43 44(A) 45 46 47 48 49 50 84 (A)' 87 102 (A) NE NE NE 9 26 (A) 28 (A) 29(A) 30 (A) 31 (A) 32 (A) 33 (A) 34 (A) 35 (A) 36 37 (A) 38 (A)

7 4 7 6 1 8 2 31 44 5 5

U 5-41 A 183a U 9-41 H 17b A 20a Su 1242

-

-

4 1

1 16 -

10 49 -

4 4 9 -

9 11 -

11 -

9 30 2 9 21 -

19 -

20 21 51 12 -

-

12 -

19 -

-

19 -

-

Ap 320c U 14-41 K 15

781-55 U 4-41 Bi 7457-41 Su 65-42 A 103 U 1-41 Bi 7458-41 Su 4344-41 F 8316-41 PA 236 A 12b 2147-59 Bi 7509-41 Pus 3432-41 G 3404-41 Bi 7575-41 E 56b A 51d A 433a A 295b A 195a A 168a A 169a A 236a A 282a A 290a A 180a PA 80c H 308b D 227 = G 7, K886CB10-1 H 330b U lla-44 C 218-70 Bi 316-42 Bi 449-42 K 14a Bi 161-42 E 69 Su 3973-41 H 36 Ap 289 E 75 A 104a A 198a A 84a A 262a

Isolated from Hu U HuA HuU Hu F Hu A Hu B Hu A Hu U Ca S Hu D HuU Hu U Hu Pus Hu A Hu U Hu U Hu U Hu F Hu P Hu A Hu F Hu U Hu P Hu B Hu U Hu P HuA HuA Hu A HuA HuA HuA HuA Hu A HuA HuA Hu P Hu F Sw D Hu F Hu F HuU Hu S Hu P Hu P Hu F Hu P Hu P Hu U Hu F Hu A Hu P Hu A Hu A HuA Hu A

References 128 296 128 138 296 128 138 128 183 24 128 128 128 296 128 128 128 128 296 39 128 128 128 128 128 138 296 296 296 296 296 296 296 296 296 296 296 138 211 *g

296 296 146 128 128 138 128 138 128 138 296 138 296 296 296 296

designation L L L B L L NE L NE NE L L L L L L L L L L L L L L B A A A A A A A A A A A A B NE NE NE L A A A A A A A A A A A A

671


VOL. 41, 1977

TABLz 1-Continued 0

K

H

Culture no.

9 9 9 9 10 11 11 11 12 13 14h 15 15 15 15 16 16 17 18ab 18ac l9ab 20 20 20 20 21 22 23 23 23 24 25 25 26 26 26 27 28 28 29 30 32 33 34 35 36 37 38 38 39 40 41 42 43 44 45 45

39(A) 55 57 NE 5 10 NE NE 5 11 7 14 NE NE NE 1 NE 16 (76) (77) NE 17 83 84 101 20 13 18 (21)' 22 + 19 23 (60) (60) (60)

9

A 121a N 24c H509d A 18d Bi 8337-41 Bi 623-42 K181 C 2187-69 Bi626-42 Su4321-41 Su4411-41 F7902-41 P 12b N234 K50 F 11119-41 P4 K 12a F 10018-41 D-M 3219-54 F 8188-41 P7a CDC 134-51 CDC 2292-55 1473 E l9a E 14a E 39a H 38 H 67 E 41a E 47a H 54 H 311b F 41 5306-56 F 9884-41 K la Kattwijk Su 4338-41 P 2a P 6a E40 H 304 E 77a H 502a H 510c F 11621-41 N 157 H7 H 316 H 710c P lla Bi7455-41 H 702c H 61 K 42

-

(73) -

NE -

74 1 NE

-

32 19 4 10 33 52 -

11 -

4 17 25 27 -

48 18 14 7 7 26 26 1 15 15 15 12 1 46 10 19 10 10 9 10 26 30 -

4 40 37 2 18 10 23

Isolated from"

HuA Hu NA HuF Hu A HuP Hu P Ca S HuS Hu P Hu P Hu U Hu F HuF Ca F Ca S HuF Sw D HuF Hu F Hu D HuF HuF Hu D Hu D

References 296 296 138 296 128 128 183 146 128 128 128 128 128 183 183 128 201 128 128 48 128 128 287 287

SwD

*.

HuP Hu P HuP HuF Hu F Hu P Hu P Hu F Hu F Hu D Ch F Hu F HuF Hu D Hu U Hu F Hu F HuP Hu F Hu P Hu F Hu F Hu F Ca F Hu F HuF HuF HuF Hu U Hu F Hu F Ca S

138 138 138 138 138 138 138 138 138 181 22 128 128 50 128 128 128 138 138 138 138 138 138 183 138 138 138 128 128 138 138 183

Former K designation A A B NE L L NE NE L L L L NE NE NE L NE L B B B L B B L L L L L L L B B B B -

-

-

NE -

-

-

L L NE

672

BACTERIOL. REV.

ORSKOV ET AL.

TABLE 1-Continued 0

46 48 49 50 51 51 52 52 53 54 55 55 56 57 58 59 60 61 62 63 64 65 66 68 69 70 71 73 73 74 75 75 76 77 78 79 80 81 82 83 83 84 85 86 86 86 86 86 86 87 88 89 90 91 92 95 96 97 98

K + NE NE -

(59) (59) + -

92 -

95 100 -

H 16 12 4 24 24 10 45 3 2 6 -

27

19 33 19 30 -

25 4 38 42 12 31 34 39 5 5 8

96 (80)

-

-

40 26 -

97 -

24 -

-

-

31 31 21

-

1

-

25

(61) 2(62) (64) (61) NE -

+

-

-

2 36 34 47 12 25 16 -

33 33 19 -

8

Culture no.

Isolated from

References

P lc U 8-41 U 12-41 U 18-41 U 19-41 K 72 U 20-41 4106-54 Bi 7327-41 Su 3972-41 Su 3912-41

Hu F Hu U Hu U Hu U HuU Ca S Hu U Hu D Hu U Hu U Hu Pus Hu D Hu Mes Hu F Hu F Hu F HuF Hu F Hu F Hu F Hu F HuF Hu F Hu F HuF Hu F Hu F HuF Hu D Hu P Hu P Hu D Hu P Hu P Hu P Hu P Hu P Hu F Hu F HuF Hu F HuF HuF Hu F Hu D HuD Hu D Hu D Hu D Hu F Hu F Hu F Hu F Hu F Hu F HuF Hu F Hu F Hu F

128 128 128 128 128 183 128 24 128 128 128 133 128 128 128 128 128 128 128 128 128 128 128 128 128 24, 128 128 128 39 138 138, 205 205 138 138, 205 138 138 138 138, 205 138 138 138 138 138 138 184 184 47 184

Aberdeen 1064 Su 3684-41 F 8198-41 F 8962-41 F 9095-41 F 10167a-41 F 10167b-41 F 10524-41 F 10598-41 K 6b K la Pla P 7d P9b P 9c P lOa P12a 6181-66 E 3a E 3b F 147 E 5d E 10 E 38 E 49 E 71 H5 H 14 H 17a H45 H 19 H23 H 35 E 990 F 1961 5017-53 BP 12665 1755-58 H 40 H 53 H 68 H 77 H 307b H 308a H 311a H 319 H 320a H 501d

44 138 138 138 138 138 138 138 138

138 138

deFi Kation -

-

-

NE B B -

-

-

-

-

L -

B -

L B L B B B -

-

-

-

VOL. 41, 1977

0

AND K ANTIGENS OF E. COLI

673

TABLz 1-Continued 0 99

100 101 101 101 102 103 104 105 106 107 108 109 110 111 112ab 112ac 113 114 115 116 117 118 119 120 121 123 124 125ab 125ac 126 127a 127ab 128 129 130 131 132 133 134 135 136 137 138 139 139 140 141 141 142 143 144 145 146 147 147 148 148a 149 150

K

H

Culture no.

99(L) 103 (A) + 98

33 2 33 8 8 12 8 33 27 10 19 39

H 504c H 509a H 510a B 41 8CE275-6 H 511 H 515b H 519 H 520b H 521a H 705 H 708b H 709c H 711c Stoke W 1411-50 Guanabara (1685) 6182-50 26w (= K10 = HW36) 27w 28w 30w 31w 34w 35w 39w 43w Ew 227 Canioni Ew 2129-54 E 611 4932-53 2160-53 Cigleris Seeliger 178-54 Ew 4866-53 S 239 N 87 N 282 4370-53 Coli Pecs 1111-55 RVC 1787 CDC 62-57 CDC 63-57

-

(58) (68) (66) (75) (90) -

+ 98 (69) + -

(72) (70) (70) (71) (63) (65) (67) -

+ -

(78) (79) (81) 12 (82)

-

18 -

21 32 18 10 4 -

27 6 10 16 30 19 6 2 -

4 2 11 9 26 28 29 35 -

41 -

(85) (85) 88ab (L) (86)

1 56 43 4 4 6

-

-

-

-

-

-

-

21 19 19 28 53 10 6

(89) 88ac(L) (89) -

NE (91) 93

SN3N/1 CDC 149-51 RVC 2907 E 68 C 771 4608-58 1624-56 E 1385 (3) CDC 2950-54 G 1253 D 357 = G 1253 K88E 519-66 E 480-68 D 616 = CS 1483 K881935

Isolated from" Hu F Hu F HuF Ca D HuF HuF Hu F HuF Hu F HuF Hu F Hu F Hu F Hu F HuD Hu D Hu D Hu F Ca S Ca S Ca S Ca S Ca S Ca S Ca S Ca S Ca S Hu D Hu D Hu F Hu D Hu D Hu F Hu D Hu D Hu? Ca S Ca F Ca F Hu F Hu D Hu D Ca D Sw D Sw D Hu F Hu D Sw D Sw D Hu D Hu D Hu D Hu D Hu D Sw D Sw D Hu D Hu D Sw D Ch S

References 138 138 138 212 * 138 138 138 138 138

138, 205 138 138 138 133 44 44 39 182, 183 182 182 182 182 182 182 182 182 42 289 44 289 47 39 290 257 50 183 183 183 50 185 185 24, 185 49 49 209 24 195 211 195 39 39 -

39 213 213 193 207 214 60

Formers K -

-

-

B B B B B B B

B

B B B B B B B -

-

-

-

B L B B B B, L B B, L B NE B -

674

ORSKOV ET AL.

BACTERIOL. REV.

TABLE 1-Continued K

0

-

151 152 153 154 155 156 157 158 159 160 161 162 163 164

-

94 88ac(L) -

-

H

Culture no.

Isolated from

10 7 4 9 47 19 23 20 34 54 10 19

880-67 1184-68 14097 E 1541-68 E 1529-68 E 1585-68 A2 E 1020-72 E 2476-72 E 110-69 E 223-69 10B1/1 SN3B/1 145/46

Hu D

-

Reference

193 193 193 206 206 206 60 207 207 207 207 209 209 237b

Hu D Hu D Hu F Hu F Hu F Sw D Hu D Hu D Hu D Hu D Hu F Hu F Hu D

Former K L -

-

-

a Abbreviations: Hu, human being; Ca, calf; Sw, swine (piglet); Ch, chicken; U, urine; A, appendix (appendicitis); NA, normal appendix; S, blood (septicemia); P, peritoneum (most often appendicitic peritonitis); B, bile; F, feces from healthy individual; D, diarrhea; Mes, mesenterial lymph node. b Bold-faced numbers = test (reference) antigens. c Bold-faced numbers in parentheses = former reference antigens, now proposed to be deleted. These numbers should not be used for new K antigens. d NE, Presence of K antigen not examined. - in K column = no K antigen detected; + in K column = polysaccharide K antigen not numbered; - in H column = nonmotile. The following polysaccharide K antigens have been found to be closely related or identical (260): K2ab - K2ac K62, K7 = K56, K12 K82, K13 K23, K18 - K22, K16 K37 K97, K53 K93, K54 K96. The scheme only contains the serotype formulas of the reference strains used at the World Health Organization Collaborative Center for Reference and Research on Escherichia and thus comprises all hitherto officially establishedEscherichia antigens. It is evident that nothing about prevalence of single antigens can be deduced from the data found in this scheme. The strains have been collected over more than 30 years, and many different considerations have determined the selection. e The famous "K-12" strain used by molecular biologists and biochemists should not be mistaken for the test strain for K antigen K12 (Su65-42). f This strain is the previous test strain of O-antigen 093 = 093:?:H- (see text) with K antigen related to K84. ° *, This review. h 014 contains no S-LPS but R-LPS. The K21 antigen has now been lost. Test strain of K24 was previously assigned to O-group 22. -

-

-

-

-

B antigen. There are also reports on cross-reactions between E. coli 0 antigens and surface antigens of mammalian cells (29, 100, 103, 272). SEROLOGY OF K ANTIGENS In 1945, Kauffmann and Vahlne (135) introduced the term K antigen (from the German word for capsule, Kapsel) as a symbol to denote either envelope or capsule antigens, of which three different types had been described, the L, A, and B antigens. The original description of K antigens was based solely on the bacterial agglutination reaction, and the inagglutinability of nonheated bacteria in 0 antiserum was considered to be the primary criterion of the presence of a K antigen (127, 128, 132, 138, 296). Classical L A B Subdivision in the Light of Recent Findings Since the K antigens (L, A, and B) were first group

-

-

defined, more than 30 years have passed, and it is only timely to consider whether any adjustments or revisions are required. Already Kauffmann and Vahlne (135, 296) pointed out the difficulties that might be encountered in distinguishing between L and B antigens. On several occasions we have run into such difficulties, either in the laboratory in Copenhagen or when evaluating work carried out in other laboratories. In the next portion, some of these problems will be mentioned. It was found that the same E. coli strain could contain two K antigens, e.g., one labeled B (K87) and the other labeled L (K88) (211). Furthermore, repeated examinations of the same strain would at one time label a strain as L and the next time as B. In the classical "L or B experiment" (Table 6), in which an OK serum was absorbed by homologous boiled culture, the outcome was dependent on variable factors, e.g., the amount of boiled bacteria used for

VOL. 41, 1977

0


675

TABLE 2. Cross-reactions between E. coli 0-antigen test strainsa O serum O antigens [fo 1 2, 10, 14, 50, 53, 107, 115, 117, 148, 149, 150, 154 2 1, 50, 53, 74, 117 3 13, 23, 53, 115 4 12, 13, 16, 18, 19, 102 5 7, 65, 70, 71, 114 6 57 7 5, 19, 25, 36, 71, 116, 141 8 32, 46, 60 9 10 11 125 12 4, 15, 16, 123 13 3, 16, 18, 19, 50, 62, 69, 129, 133, 147 14 24 15 12, 40, 45, 143 16 4, 46, 129, 135 17 18, 44, 62, 73, 77, 106 18 4, 13, 16, 19, 23, 133, 138, 147 19 13, 33, 39, 133, 147 20 21 22, 32, 83 23 3, 13, 18, 38, 68 24 14, 56 25 4, 7, 13, 18, 19, 26, 36, 68, 102, 133, 138, 147, 158 26 4, 13, 25, 32, 100, 102 27 28 29 30 32 8, 21, 26, 83 33 34 4, 85, 140 35 36 25, 43, 109 37 48 38 23 39 7, 91 40 15 41 42 43 36, 118 44 68, 73, 77, 106 45 15, 54, 66 46 8, 16, 134 48 19, 54, 59 49 50 1, 2, 13, 19, 44, 53, 107, 117 I 133, 135

51 52 53 54 55 56 57 58 59 60

1, 2, 3, 50, 149 45, 48, 59

0 antigens

serum

61 62 63 64 65

5, 70, 71

66

45

68 69 70 71 73 74 75 76 77 78 79 81 82 83 84 85 86 87 88 89 90 91 92 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110

[fo serum

108

119 120 121 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163

13, 16, 17, 40, 68, 73, 106 154

7, 13, 18, 25, 36, 44, 62, 102 13, 51, 150 5, 65, 70, 74, 116 5, 7, 65, 70 13, 17, 44, 56,62,68,77, 106 2, 40 1, 163 22 17, 44, 66, 73 92, 116, 137 41

37, 51 21, 22, 32, 46 34, 140 19, 48, 90, 127 41, 48, 76, 116 141 115

19, 86, 127 39

19, 78, 91

13, 63 26 117, 162 4, 25, 26, 36

17, 44, 62, 73 50, 102, 117, 123 61

O antigens

3, 48 53, 102, 105, 115, 117 101, 116, 123 12, 116, 121 11, 73

86, 90, 128

13, 16, 133, 135 113, 125, 143 13, 129, 135, 147 46

13, 16, 17, 50, 129, 133 78

7, 18, 25, 148, 150

102 34 7, 88

3, 15, 131

112, 149 13

19, 102, 133 1, 138 1, 50, 53, 112, 144 1, 69, 138

3, 115 1, 64 7 25

101 75

111

112 113 114 115 116 117 118

144, 149 112, 117, 131 5, 48 1, 3, 152 7, 123 50, 76, 101, 122 121, 123

24 6

48, 54 8

L

The data in this table have been collected

L. mainly

from examinations

through

many years in the

laboratory

in

Copenhagen. However, some of the data of Edwards and Ewing (39) are included, especially those reactions that have been detected in both laboratories, even if titers may have been low. Several strong reactions found in their laboratory but not in Copenhagen are also listed. Titers are not recorded, as it is our experience that titers of cross-reacting antigens may vary greatly among sera produced at different times. The reactions recorded are generally those that have been found in serum dilutions not less than five twofold titer steps below the homologous titer. The papers by Kampelmacher (122) and Glantz (69), who have made similar studies, should also be consulted by those who are interested in cross-reacting E. coli 0 antigens.

BACTRIUOL. RzV.

ORSKOV ET AL.

676

TABLE 3. Cross-reactions between 0 antigens of E. coli from dysentery-like disease and Shigella (according to Edwards and Ewing [390a E. coli Shigella S. boydii 13 028ac (Kattwijk) 0112ac (Guanabara) S. dysenteriae 2, identical 0124 S. dysenteriae 3, identical 0136 0143 S. boydii 8, identical 0144 S. dysenteriae 10 0152 Not examined a Other strong 0-antigen cross-reactions between E. coli and Shigella are: 032/S. boydii 14, 053/S. boydii 4, 058/S. dysenteriae 5, 079/S. boydii 5, 087ab/S. boydii 2, 0105ab/S. boydii 11, 0112ab/S. boydii 15, 0129/S. flexneri 5, and 055/S. flexneri 4b. TABLE 4. Cross-reactions between 0 antigens of E. coli and Salmonella based on the results of Kauffmann (130), Frantzen (57), and Refai and Rohde (229) E. coli

Salmonella 042,

01

055 040,403 059 038 051

02 06 015 021 023 044, 062, 068 070, 073, 099 0106 and 0129 055 075 085 086, 090 0111 0132 0134

06, 14 0501502504 011

017 043 035 017 036

TABLE 5. Cross-reactions between E. coli and K. pneumoniae 0 antigens E.

coli

Ol9ab 09 020 08

K.

01, 03, 04, 05,

pneumoniae

identical to 019b identical strong relationship identical

absorption. Very often, most, if not all, antibodies against the polysaccharide K antigen were removed by the standard absorption and, if not, repeated absorptions would remove the rest. The antibodies of a so-called pure L serum were therefore either a mixture of residual antibodies against the polysaccharide K antigen and other thermolabile surface antigens or simply antibodies against such thermolabile structures (fimbriae, fimbria-like antigens, H antigens, or other, undefined structures). If no detectable antibodies were left against thermolabile structures or against the aqcidic polysaccha-

TABLE 6. Schematic presentation of the agglutination results on which previously used definitions of K antigens (L, B, and A) are based OK serum K

type

Antigen prepn

0 serum

Absorbed

by culture heated at 100°C for 2

UnabOrbed

aob-d

h L

B

Live (or Formalin treated) Boiled (100TC for 1 h) Live Boiled

+b

+

+

-c

+

-*

-

+ +

a

+

+ Live + Boiled a *, Negative or significantly lower than that of boiled culture. b +, Agglutination. e -, No agglutination.

A

ride K antigen, the strain was labeled B. The above-mentioned facts were not recognized when the L and B definitions were made originally. As the quantitative development of the different surface structures can vary much from time to time and from laboratory to laboratory, dependent on variations in growth conditions, the observed variable outcome ofthe absorption experiment is easy to understand. Thus, the agglutination experiment, although it can be helpful in the preliminary examination, cannot be relied upon as the sole test for antigenic analysis. Methods that can separate the single antigen-antibody reaction must be introduced. At present, the agar precipitation techniquhes are the best methods for this purpose. By use of such serological techniques, three representative E. coli K antigens of the L variety, K12, K51, and K52, were reexamined by Orskov and 0rskov (203, 204). In all three strains the presence of heat-stable, extractable K antigens was demonstrated by passive hemagglutination, double diffusion in gel, and immunoelectrophoresis. In the last test they showed a high electrophoretic mobility toward the anode. The three K antigens are acidic polysaccharides; their sugar compositions or structures have not yet been examined. Similarly, the K2ac antigen, originally described as an L antigen, was studied by Holmgren et al. (101, 102). The antigen was found to be a polysaccharide of high electrophoretic mobility, and it retained its precipitating as well as its agglutinin-fixing capacity after boiling. Later, immunoelectrophoretic studies in agar of all E. coli K-antigenic test strains have shown anodic precipitation arcs in extracts heated at 100'C due to K antigens in all former

VOL. 41, 1977

A- and L-antigen-containing strains, except 0137:K79(L) (194). From 33 strains, in which B antigens had been described, the presence of a K precipitation arc could only be demonstrated in a few cases. These were the three first-established B antigens, K25, K56, and K57, and in addition K82, K83, K84, and K87. When Knipschildt described the three first B antigens, he was able to produce pure B antisera by absorption with other strains of different K types. However, when the next B antigens were numbered in strains from infantile diarrhea [O111:K58(B) and 055:K59(B)], this was done with great hesitation because there was uncertainty as to the existence of a separate K antigen, since it was not possible to produce a pure B antiserum by absorption (133). The presence of B antigens in these cases and with a few exceptions in all later cases was thus based solely on the inagglutinability of the live culture in an 0 antiserum. Thus, although it is highly probable that such strains do not contain an acidic polysaccharide K antigen, it is important to remember that antisera produced with a live culture (OK sera) are preferable, and often necessary, for their primary detection by slide agglutination. Several substances and structures can cause inagglutinability in 0 serum, and often it will not be possible to point out which factor(s) is responsible in a certain case. The most important factor that interferes with agglutination in O antiserum is the acidic polysaccharide K antigen. However, many strains are found that are agglutinable in 0 serum, although they are equipped with an acidic polysaccharide capsule, as shown by immunoelectrophoresis (190) or agar electrophoresis combined with Cetavlon precipitation in the second dimension (187). H antigens (flagella), fimbriae, and perhaps other surface structures can cause inagglutinability in 0 serum. Furthermore, it is a common experience that this inagglutinability can be abolished by changes in the growth medium or the growth temperature. Another common experience is that strains, originally inagglutinable in 0 serum, become agglutinable after passages in the usual media. This applies to many of the test strains for the former B antigens found in so-called enteropathogenic strains, e.g., 055:K60(B6) and O111:K58(B4). In these cases it is not possible to detect any serological differences between such laboratory strains that are agglutinable in 0 serum and freshly isolated strains of the same serotype that are inagglutinable in 0 serum. It should be emphasized that we do not know which factor(s) causes this inagglutinability of the freshly isolated strains, but we believe that the variable


677

agglutinability in 0 serum alone does not warrant the description of a special K antigen. In light of the above considerations, we find it misleading to continue labeling K antigens according to the classical L, A, and B criteria and shall therefore propose to restrict the nomenclature of K antigens to (acidic) polysaccharide K antigens and protein (fimbrial) K antigens. The polysaccharide K antigens may be subdivided into two groups: those found in combination with 08, 09, 020, and 0101 and those found in probably all other combinations (see below). However, until more knowledge has been gained, we shall abstain from giving any special name to these subgroups. Several capsulated strains belonging to 08, 09, or 0101 are equipped with capsules that make the bacteria inagglutinable in 0 serum even after boiling, but heating at 1200C for 2 h will make them agglutinable. The K antigens of such strains will, for practical reasons, still be denoted K(A). It should be stressed that these strains can lose this special heat resistance of the capsule by mutation, and thus it is not uncommon in one animal to find strains of the same OK serotype with and without the special heat resistance of the K antigen. Polysaccharide K Antigens Techniques for K-antigen determination. The bacterial agglutination technique used for the determination of E. coli K antigens is performed in tubes or, preferably, on slides with 0 and OK antisera produced with heated and nonheated cultures, respectively; inagglutinability in 0 serum and agglutinability of a live culture in OK antiserum can be taken as an indication of the presence of a K antigen. For slide agglutination, the culture from a plate is suspended directly in a drop of antiserum, whereas a suspension of the culture in saline-formaldehyde is used for tube agglutination. The tubes are incubated for 2 h at 370C, left at room temperature for about 20 h, and then read. The agglutination technique, which is fast and sensitive, gives an idea of the combined agglutinating abilities of the different antigens on the bacterial surface, but often little information concerning the role played by the single antigens. Strains subcultured exclusively on solid media may be both flagellated and fimbriated. OK antisera raised with such bacteria will therefore contain antibodies against these structures, and there are probably additional surface components that might cause erroneous results. Agglutination reactions due to these antigen-antibody systems have often been interpreted as being due to thermolabile K anti-

678

ORSKOV ET AL.

gens (L antigens). Thus, a description of an E. coli K antigen should not be based solely on the agglutination reaction, but also on gel diffusion techniques such as double diffusion in gel (218, 219) or immunoelectrophoresis (241), by which the K antigen is directly recognizable. Slide agglutination in single or pooled K (OK) antisera may be used as a preliminary test, but in most cases the method of choice is the countercurrent technique (18, 76, 260), by which the result is read after a 1-h electrophoretic run. During this time the homologous antibody migrating towards the cathode and the K antigen migrating towards the anode will meet and give a precipitation line. The serum agar technique of Bradshaw et al. (13) has been adapted for K determination of a limited number of K antigens by Kaijser (121a). Antigen extracts for the diffusion technique can be made in several ways. Holmgren et al. (102) suspend acetone-dried bacteria in Veronal buffer, pH 8.6, at 370C for 24 h. The supernatant of the suspension after centrifugation is used as the antigen (VE extract). A freeze-press extract can be prepared as described by Edebo (36). Holmgren et al. (102) consider the VE extract to be the most representative antigenic preparation containing the largest number of demonstrable precipitinogens. Glynn and Howard (71) use both crude (homogenized acetonedried culture) and purified (by precipitation with Cetavlon) extracts for immunoelectrophoresis. The extracts used in Orskov's laboratory for immunoelectrophoresis (194) are produced by suspending a culture from agar plates in buffered saline, pH 7.3, and heating at 60'C for 20 min, followed by centrifugation. The supernatant is called the 600C extract. Part of this is further heated at 1000C for 1 h and called the 60/100'C extract; this extract is useful if only 0 and K precipitation lines are wanted. The same kind of extracts can be used for the different electrophoretic methods. For countercurrent techniques the extract has to be diluted (e.g., 1:200), or a simple preparation can be used, i.e., a suspension heated directly at 1000C for 1 h without centrifugation. For the evaluation of the amount of K antigen, Glynn and Howard (71) use the ability of this antigen to inhibit the agglutination of sheep erythrocytes by rabbit antibody, a method described previously with pneumococcal polysaccharides by Ceppellini and Landy (20). Kaijser (121) finds, however, that the amount of K antigen is more specifically measured by crossed immunoelectrophoresis. Test K antigens. So far, K antigens numbered K1 to K100 have been established (Table

BACTERIOL. REV.

1). According to our present knowledge, only K1 to K57, K62, K74, K82 to K84, K87, K92 to K98, and K100 represent acidic polysaccharide K antigens. K88 and K99 are proteins. The remaining earlier established K numbers, many of which are so-called B antigens, have not been demonstrated as special K antigens independent of the 0 antigens. The patterns obtained with all hitherto established K test strains in immunoelectrophoresis in agar are shown in Fig. 1, which is a revised combination of similar figures presented before (190, 194). Previously, an immunoelectrophoretic group was named lAb because antibodies against the weakly immunogenic K1 and K5 are difficult to demonstrate. Strains of IE group lAa are now combined with those of lAb in a common group 1A. IE group lBa only contains K antigen strains having 0 antigens 08, 09, 020, and 0101. The K antigens of this group generally do not move far away from the application basin, as opposed to the K antigens of IE group 1A and, furthermore, the 08, 09, 020, and 0101 precipitation arcs are usually weak or even missing in strains giving strong K-antigen precipitation arcs (see reference 190). Many strains listed in IE groups lBb and 2b as K-antigen test strains are disregarded as K strains today because in these strains K antigens of the B type were only defined by the inagglutinability of a nonheated culture in 0 antisera. K95 to K98 and K100 are the latest reported K antigens (205). K95 to K98 were found in previously established 0 test strains by immunoelectrophoresis examination. Some additional K antigens were found in other 0 test strains by this method, but they will not be numbered, mainly because the reactions are difficult to reproduce and too erratic. These strains are denoted K+. The K antigen in 045 is denoted K1, as it is closely related or identical to test antigen K1. Form variation in E. coli polysaccharide antigens was described by 0rskov et al. (194). Similar to the form variation observed for Salmonella 0 antigens (132), E. coli strains with the K1 capsular polysaccharide can be found in two morphologically identical colony types designated K1+ and K1-, based upon agglutination or immunoprecipitation techniques. The variation phenomenon occurs in various K1 strains with a frequency that ranges from 1:20 to 1:50. In some cases, strains can be found that are stable in the K1- form. It is possible to produce a specific K1+ antiserum by absorption of an OK K1+ serum by a K1- stable strain. Although the K1- form is hardly immunogenic,

VOL. 41, 1977

injection of the whole bacteria in the K1+ form usually gives low-titered antiserum. Our recent studies have shown that these antigenic differences can be related to structural differences in the K1+ and K1- form variants. Both K1+ and K1--purified polysaccharides are homopolymers of alpha-2,8-linked polysialic acid. The K1+ polysaccharide has approximately 80 to 90% of the sialic acid residues 0acetylated; in contrast, the K1- variant has little or no detectable O-acetyl groups (F. (rskov, I. 0rskov, R. Schneerson, W. Egan, A. Sutton, and J. B. Robbins, manuscript in preparation). Recently, three new K antigens have been serologically examined in strains received for K-antigen determination. These are E. coli 1473 (020K101:H-), E. coli 6CB10/1 (08:K102:H-), and E. coli 8CE275 (0101:K103:H-). Strain 1473, which was obtained from H. W. Moon, Ames, Iowa, causes diarrheal disease in newborn, colostrum-deprived piglets. The two other strains were isolated by K. A. Bettelheim and S. M. S. Lennox-King, London, in a maternity ward. The K84 antigen deserves attention. The test strain of this antigen has the 020 antigen. However, a K antigen strongly related to K84 can also be found in combination with 08 as in strain H308b. That strain was until recently the test strain for 0 antigen 093, and it was accepted that 093 and 08 were cross-reacting, independent 0 antigens. With the aid of the 0specific phage M8, it was shown (114, 302) that 0 test strain 093 was in fact an 08 strain with a K antigen. The same result was obtained independently by immunoelectrophoretic studies. Thus, the supposed 0 antibodies in the 093 "0 serum" were actually a mixture of antibodies against a heat-stable K(A) antigen and an 08 antigen, overlooked thiough many years. Cros-reactions between E. coli K antigens. Many E. coli strains will agglutinate on slides in a great number of E. coli OK antisera, and many kinds of surface antigens may participate in these reactions. Many fewer cross-reactions are found by a gel precipitation technique, such as double diffusion in gel or countercurrent immunoelectrophoresis. Cross-reactions expressing relationships between K antigens have been found among the following K antigens: K18-K22-K100, K13-K20-K23, K53-K93, K54K96, K16-K97, K37-K97, K12-K82, K2ab-K2acK62, and K7-K56. K2 (ab and ac) is so closely related to K62, K7 to K56, and K12 to K82 that K62, K56, and K82 will probably be deleted as the K antigen in favor of K2, K7, and K12, respectively. The other cross-reactions mentioned need further examination.


679

Cross-reactions between E. coli 0 antigens and K antigens. Identity between the unnumbered 0 antigen in strain 145 and the polysaccharide K antigen K87 found in strain G7 (08:K87, K88:H19) has been described (211). Furthermore, an antiserum raised with a boiled culture grown at 370C of strain G7 contains both agglutinins and precipitins against the 08 antigen but not against the antigen having the K87 specificity. However, an 0 antiserum prepared with a boiled culture of G7 grown at 180C does contain agglutinins against the K87 antigen. The test strain of K87 was formerly strain 145, but is now a K88- mutant of G7 called D227 (08:K87:H19). Examples of relationships, but not of identity, between 0 and K antigens are the crossreaction between the following: K31 and 0120 (205), K9 and 0104, K44 and 053, and K45 and 074 (unpublished data). Cross-reactions between E. colU K antigens and polysaccharide K antigens outside the Eswherichia group. Morch and Knipschildt (170) described cross-reactions between some E. coli strains having K(A) antigens and the capsule antigens 9 and 23 of Streptococcus pneumoniae. Unfortunately, these strains have not been kept. Cross-reactions between K antigens of E. coli and Klebsiella have not yet been published. Heidelberger et al. (91) described antigen relations between E. coli K30 polysaccharide and pneumococcal polysaccharides Sn II and SV andE. coli K42 and Sn XXV, as well as K85 and S I and SV. Grados and Ewing (75) and Kasper et al. (124) demonstrated crossreactions between E. coli K1 and meningococcal group B polysaccharide. Bradshaw et al. (13) used agar plates containing anti-Haemophilus influenzae type b serum to discover bacteria producing cross-reacting antigens. Schneerson et al. (252) described an E. coli strain with a polysaccharide K antigen closely related to Haemophilus type b capsule antigen; the strain was first described as 075:K(F147):H5 and is now test strain 075:KIOO:H5 for the new K100 antigen (205). Cross-reactions between the capsular antigen of Neiseria meningitidis group C and E. coli K antigen of strain Bos 12 and other strains were described by Robbins et al. (232). This K antigen was later designated K92; Bos 12 is now 016:K92:H-. Cross-reactions between S. pneumoniae type 3 capsule and E. coli K7 antigen were described by Robbins et al. (234). The immunological implications of these and similar cross-reactions are at present intensively investigated in many laboratories. Robbins et al. (234) have recently reviewed and discussed the relevant literature.

IMMUNOELECTROPHORETIC PArrERNS OF E. COLI O AND K ANrIGEN rESr STRAINS 01:K1 01 :K51 02:K1( 02:1K56 03: K2 04: K3 04: K6 04: K12 04 1K52 05;.14 069:K2 06:1(3 06: K15 06::K53 06:K154 K10 012: KS 07:K1( 07:1K7 010: KS 0117 Q01: K11 015: K14 016: K1 017: K16 021: K20 023: K1( 023: K22 025: K19 025:1K23

+ o

1A

o_>

_,

_

_K -n

044: K74 045: K1 049: K * 073: K92 075s K95 075: K100 077:196 086: K62 0103:K1( 0107: K98 0117:1(98 0154 : 194

. 0

08: K8 08: K25 08: K27 0:1K40 00: K41 08: K42 08: K43 08: K44 08:&45 08 :K6 08:1K47 08:1K48 08: K49 08: K50 08: K87 08: K102

z

09:1K29 .K. Q9 9 09:1K26 09:1K28 09:K33 09:Q 09: ff3 09: ______________ fL4:

OK

09: K35 09 : K36 09: K37 09 09: K38 09: K39 09: K55 09: K57 020:K17 020:K83 020:1K84 020: K101 0101: K103

1B_

018::K76 018::K77 019:K- 023 :K21 026:K60 035:K- 036:K- 039:K- 040:K- 043:-K 050:K- 051:K- 052:K- 055:K59 060:K062:K- 063:K- 066:K- 068:K - 070:K071:K- 073:K- 078:K80 085:K- 086:K086: K61 086: K64 034: K -

b.

O

B

08:K- 089:K- 090:K- 092:K-

B .

0

-

0

21 b

ON o

,

095:K- 097:K- 099:K- 0101:K0102:K- 0106:K- 0109:K- 0111:K58 0114:1K90 0118:K- 019:1K69 0123:K0125:K70 0126:K71 0127:K63 0127: K65 0128::K67 0129:K- 0132:K0135: K- 0142:1(86 0145:K- 0148:K0151: K- 0153:K- 0155:K- 0156:K0157: K- 0162: K

014:1K7 022: K13 081:1K97 083: K24 0116:1K 0120:1K 0139: K82 0150: K93 027:K- 028ab:K- 028ac:1K73 029:K030:K- 032:K- 037: K- 038:1K 033:K041:K- 042:K- 046:K- 048:K- 053:K054: K- 057: K- 058: K- 059: K- 061 :K064:K- 065:K- 069:K- 074: K- 076: K079:K- 080: K- 02: K- 083:K- 084: K087:K1- 091:1K- 096: K- 098: K- 0100:K10104:K- 0105:K- 0108:K- 0110;K- 0172ab:K68 0112ac: K66 0113: K75 0115: K - 0121: K 0124: K72 0130: K - 0131: K - 0133: K 0134: K - 0736:K78 0137: K79 0138:1(87 0140:K - 0141: K85 0143: K0-044: K 0146:K- 0147:1K89 0149:1K9 0152:K- 0158:K- 0159:K0160:K- 0161:K- 0163:K-

680

VOL. 41, 1977

M Antigen Many Enterobacteriaceae strains produce a slime (mucus) antigen that has been termed M antigen (125, 126). Although it forms capsules, it is nonspecific and thus distinct from the K antigens discussed here. It has been suggested by Kauffmann (132) and Henriksen (95, 96) that the M antigen, regardless of the microorganism that elaborates it, in all instances is serologically the same or nearly the same. The M antigen that is also known as capsular antigen (163), colanic acid (73), and slime wall antigen is usually better developed at a growth temperature lower than 3700. For its formation, the composition of media is very important. It has been shown (1) that solid media with concentrations of solutes giving a high osmotic pressure will cause normally nonmucoid Enterobacteriaceae to produce great amounts of mucous substance. The M antigen can be found in many E. coli strains. The M antigen cross-reacts serologically with some type-specific K antigens, such as the K30 and K39 antigens of E. coli and the K8, K11, K13, K21, and K35 antigens of Klebsiella (97). Antisera against E. coli or Klebsiella strains of the above K types cannot always be used for demonstration of the M antigen; usually only some of the immunized rabbits will produce these cross-reacting antibodies. Protein K Antigens (Fimbriae Classified as K Antigens) One of the E. coli K antigens, K88, is known to be a protein (277). According to bacterial agglutination results and classical definitions, this antigen behaves as a typical L antigen (211). In an electron microscope it is seen as long and thin fimbriae covering the whole surface of the bacterial cell (277). The antigen is described as fimbria-like. It can also be demonstrated as a surrounding coat by the immunofluorescence technique (lrskov and Lind, unpublished data) and after staining with Leifson flagella stain (189). Another similar antigen, K99, is also a protein (105a). Both antigens are found in strains associated with diarrhea in animals, K88 mainly in swine and K99 mainly in calf and lamb strains, and both confer agglutinating ability on blood cells that cannot be in-


681

hibited by mannose, K88 on guinea pig erythrocytes (118, 277) and K99 on horse erythrocytes (291). Both are transferable by plasmids and can be found together with chromosomally determined K antigens. No cross-reactions are known to exist between K88 and K99 and other antigens. Recently a fimbrial antigen has been demonstrated in a strain from diarrhea in humans and claimed to act as a colonization factor (41). Another antigen has been found in some pig strains without the K88 antigen (105a). Furthermore, an antigen found in some calf strains probably belongs to the same class (297a). These have not yet been numbered as K antigens. Fimbriae Not Classified as K Antigens Nonflagellar filamentous appendages were first recognized in E. coli by Houwink and van Iterson (104). They were later called fimbriae (35) or pili (16). They were demonstrated in many Enterobacteriaceae (30, 32, 34, 35) and classified into several types according to morphology and adhesive properties. Some strains undergo a reversible variation between a fimbriate phase and a nonfimbriate phase, which can be influenced by the growth conditions (31). Broth culture is selective for the fimbriate phase, whereas nutrient agar is selective for the nonfimbriate phase. Type 1 fimbriae, found in mostE. coli, generally number 100 to 400 per cell in the fimbriate phase. They are characterized by a direct hemagglutination reaction and cause pellicle formation. They consist of protein (16, 180). Some strains are exceptional in that they show hemagglutinating activity but no fimbriae in an electron microscope. E. coli strains with fimbriae of type 1 give strong agglutination reactions with erythrocytes from most species, whereas nonfimbriate hemagglutinating strains have a weak activity and a more limited spectrum, and those ofdifferent strains agglutinate different erythrocytes. Type 1 fimbriae hemagglutination is inhibited by n-mannose and not affected by temperature. In contrast, nonfimbriate haemagglutinins are not inhibited by mannose and are best demonstrated in the cold at 4°C (31).

FIG. 1. Immunoelectrophoretic patterns ofE. coli 0- and K-antigen test strains. In the trough: homologous 0 or OK antiserum. In the well above the trough: 60°C/20-min extract. In the well below the trough: 60/100°C extract. The right column contains the O:K serotypes ofall E. coli 0- and K-antigen test strains. Underlined antigens = test antigens. In the test strains of 0 antigens 024 and 056, anodic lines close to the basin are seen; K antigens have been demonstrated chemically in these strains, but, since we do not know whether the lines in immunoelectrophoresis represent 0 or K lines, the strains have been omitted from the scheme. The K antigens belonging to immunolectrophoretic groups lBb and 2e most likely represent 0 specificities (LPS) and can thus be left out of the serotype formula (see text and Table 2). (Note: 0155.K- should be moved from group lBb to 2b.)

682

0RSKOV ET AL.

BACTERIOL. REV.

By raising antisera against fimbriate E. coli the bacterial cell. It consists of glucosamine and Shigella flexneri type 1 strains and testing phosphate and fatty acids, the most prominent these by cross-agglutination and absorptions, of which is ,-hydroxymyristic acid. Its strucGillies and Duguid (67) obtained results from ture has been described recently (158). Lipid A which they concluded that the fimbriae of E. is responsible for the general biological (endocoli contain a major type-specific antigen toxic) properties of the LPS (158). It will not be shared only within groups of related strains, as dealt with in this article. well as minor coli-flexneri shared antigens. Region 2 (core) is linked to lipid A via a However, no extensive systematic examination carbohydrate component that is typical for the of the serology of E. coli fimbriae has been LPS of gram-negative bacteria, 2-keto-3-deoxyundertaken. Probably several morphological as mannulosoctonic acid (KDO). Whereas there is well as antigenic types exist, not only in the E. only one structure of lipid A, found in all entercoli group, but also sometimes on a single E. obacterial LPS so far studied, five different core coli cell. oligosaccharides are described. Thus, there is a In many respects fimbrial antigens could be greater degree of variation in the synthesis of considered as protein, K antigens; they are the core oligosaccharide. The core expresses R thermolabile, superficial, and liable to confer specificity, which in wild-type S forms is cryptic relative inagglutinability in 0 serum (67). because of the substitution of the core with Other filaments that are necessary for the region 3. conjugation process and determined by conjugaRegion 3 is the O-specific polysaccharide of tive plasmids have been described. They are the LPS of bacterial S forms (S-LPS). It consists called sex pili (147), as opposed to fimbriae = of oligosaccharide repeating units. This is a common pili. Five distinct types of sex pili have structural principle that is found in all bactebeen identified and visualized by electron rial polysaccharides. Composition and strucmicroscopy in E. coli K-12. They have charac- ture of region 3 are the chemical bases of the 0 teristic morphologies and serological specifici- specificity of gram-negative bacteria. Thus, ties and are receptors for specific phages (for a with the many distinct 0 specificities in E. coli, review, see reference 23). The first type of sex many S-LPS with distinct O-specific polysacpilus found was called F because it was deter- charides are found. From the above considerations it follows that mined by the F factor (17, 147). This F pilus is the structural variability becomes increasingly identical to the f+ antigen (139, 200). Mention should be made here of previously greater in the outermost (probably phylogenetidescribed surface antigens in E. coli such as the cally younger) compartments of the LPS molea (273) and 8 (174) antigens. Fimbrial antigens cule. bear resemblance to these, but no thorough Isolation and Purification comparison has been made. The anodic thermolabile antigen of Seltmann (259) is a common Depending on the aim of research, methods antigen detected by immunoelectrophoresis were devised to isolate LPS-protein complexes, and found in all Enterobacteriaceae strains. LPS, or the polysaccharide moiety of LPS. These have been described in several reviews CHEMISTRY OF 0 ANTIGENS (112, 115-157), and only a few will be mentioned The 0 antigens have been studied exten- here. LPS may be extracted from the bacteria with sively during the past years, and more general information on them can be found in a number ethylenediaminetetraacetic acid (300) or 45% of comprehensive articles and reviews (59, 89, aqueous phenol at 65°C (303, 304). Due to their 112, 155-157, 175). They are LPS, the general high apparent molecular weights (5 x 106 to 50 structural feature of which is given in Fig. 2. x 106; see references 236 and 262), they can be The molecule consists of three regions: lipid A, purified by repeated ultracentrifugation. More an oligosaccharide termed core, and the O-spe- recently, electrodialysis was introduced as a cific polysaccharide chain shown in Fig. 2 be- purification procedure (61a). low. Region 1 (lipid A) is the part of the molecule that by hydrophobic interaction with lipoprotein (14, 15) is buried in the outer membrane of

10-specific polysaccharide - Core oligosaccharide]- Lipi

A

FIG. 2. Schematic diagram of the general structure of bacterial LPS (from Luderitz et al. [157]).

Chemical Characterization For chemical characterization, 0 antigens (S-LPS) are usually hydrolyzed and analyzed as to their sugar constituents by use of chromatographic and electrophoretic methods. This was done with more than 100 E. coli test strains, and the results were compiled into a scheme of

O AND K ANTIGENS OF E. COLI

VOL. 41, 1977

chemotypes (197). Due to the common occurrence of glucosamine (from lipid A and often from the core), glucose, galactose, heptose, and KDO (from the core), these sugar constituents were termed basal sugars (for reviews, see references 155 to 157). In many 0-specific polysaccharides (region 3) one or more of these basal sugars are missing. Wrong conclusions about the composition ofthe 0-specific polysaccharide may then be drawn from the sugar composition of the whole LPS. This explains why the sugar compositions given in Tables 7 and 9 differ from those published earlier (197). To study the 0-specific polysaccharides, LPS were degraded by mild acid hydrolysis into lipid A (region 1) and carbohydrate moiety (regions 2 and 3). Gel permeation chromatography of the carbohydrate moiety on Sephadex (106, 173, 247) gave elution patterns as shown in Fig. 3. Comparative analyses indicated that the material from peak 1 is the 0-specific polysaccharide bound to core oligosaccharide (regions 2 and 3), that from peak 2 is unsubstituted core (region 2), and that from peak 3 is predominantly KDO, which was split off during the acid degradation. The presence of unsubstituted core oligosaccharide in preparations of wild-type SLPS was an indication of the presence of R-LPS (lacking the 0-specific polysaccharide) in S forms of E. coli. More recently this was confirmed by polyacrylamide gel electrophoresis of intact LPS in the presence of sodium dodecyl sulfate (106). This method also revealed that S forms often produce LPS molecules with different chain lengths (see also reference 171).

Neutral Polysaccharide Chains The sugar composition of some neutral 0specific polysaccharide chains is given in Table 7. It is noteworthy that unusual amino sugars and also rhamnose occur quite frequently in E. coli (see also references 111 and 112). The 0specific polysaccharides may be composed of up to six different sugar constituents. There are

E C

0.8-

o 0.6 I-.

I

n

I

I'%

0.2-

I'

OK 0.2LO

50

60

90 100 110 fraction number chromatography on Seph-

70

80

FIG. 3. Gel permeation adex G-50 of degraded polysaccharides from S-LPS (-) and R-LPS (-----).

683

only two 0-specific homopolysaccharides hitherto described in E. coli, namely, the 08- and 09-specific mannans. 0-specific homopolysaccharides (mannans and galactans) are also found in Klebsiella (176). The E. coli 09 antigen is identical to the Klebsiella 03 antigen, and the E. coli 08 antigen has a structure that is very similar to that of the Klebsiella 05 antigen (see Table 8 and reference 153). The structures of the neutral 0-specific polysaccharides that were previously elucidated are given in Table 8.

Acidic Polysaccharide Chains For a long time all E. coli 0 antigens were thought to contain only neutral polysaccharide chains, comparable to Salmonella 0 antigens. More recently, LPS were isolated that contained acidic components such as glycerol phosphate (107), hexuronic acids (194; K. Jann, unpublished data), neuraminic acid (109; B. Jann, unpublished data), or 4-O-(1'-carboxyethyl)-Dglucose (glucolactilic acid [see reference 28]). Degradation studies and polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate showed that the LPS preparations, obtained in the usual way as the sediment of a preparative ultracentrifugation, contained only little polysaccharide material with acidic components. These preparations consisted largely of LPS with an unsubstituted core (R-LPS) and LPS with only one 0-specific repeating unit linked to the core (SR-LPS). An analysis. of the supernatant of preparative ultracentrifugation revealed that the major portion of the acidic LPS is not sedimented during ultracentrifugation and can be obtained from the supernatant by fractional precipitation with Cetavlon (see references 105 and 116). The sugar composition of all acidic 0-specific polysaccharides, analyzed in Freiburg until now, is given in Table 9, and the structures of some polysaccharide moieties are presented in Table 10.

CHEMISTRY OF R ANTIGENS R mutants that arise spontaneously from S forms or that may be produced in the laboratory by treatment with mutagens have LPS lacking an 0-specific polysaccharide moiety (region 3 in Fig. 2). These LPS are termed R antigens. They consist of lipid A and the core oligosaccharide, whereby the core may be complete (in rtb mutants [245, 247]) or incomplete (in rfa mutants [245, 248]). The E. coli strains K-12, B, and C are R forms of which the S ancestors are not known. E. coli K-12 is an R(rfb) mutant, and E.coli B is an R(rfa) mutant (245).

684

ORSKOV ET AL.

BACTERIOL. REV.

TABLE 7. Sugar composition of some neutral O-specific polysaccharides from E. coli

E. coli 0 antigen

,

°

X

X

0o

XX o

X

e

6

16

a

~~~~~~~~~++ ++

1

2 3 4 5 6 7 8 9 10 19 20 50 55 71 75 85 86 88 102 ill

E

+ +

+

+

+ +

+

+

+ ++

+ + +

+

+

+

+ + + +

+

+

+

+

+

o

+ +

+ +

+ + +

+

+

For the chemical study of the LPS from R mutants (R-LPS), it is preferable to extract the bacteria with phenol-chloroform-petroleum ether at room temperature (61). Thus, only RLPS, which may be considered as glycolipids, are extracted, and the purification of R-LPS is easy. R Antigens with a Complete Core From early studies it was originally concluded that all gram-negative bacteria had only one and the same core structure in their LPS, a view that holds for all Salmonella (see references 122, 156, and 157). Serological studies of Moller (169) in the late 1940s had indicated that two types of R mutants occur in E. coli strains. This was verified later by the isolation and analysis of two distinct core types from E. coli 08:K27- and 08:K42-, which were termed coli R1 and coli R2, respectively (247). More core types were found in E. coli (244-246), and their compositions are given in Table 11. Their core structures and the Salmonella core are the only complete core structures known today. The core types found in E. coli also occur in other spe-

+

+ +

+

+

+

++

+

cies, and some examples are included in Table 11. Because the coil core types also occur outside E. coli, the specification coil has been omitted from their denomination. Recently, it was shown (249) that E. coli 014 is an R strain (lacking 0 specificity) of the R4 type. The structures of some core oligosaccharides are presented in Table 12. For comparison, the Salmonella core is also shown. All core oligosaccharides seem to have a common basal structure, including the heptose units and the 1.3 1.3 adjoining sequence Glc(Gal) 3 Glc - Hep, in which an anomeric exchange of the 4-OH (Glc versus Gal) does not seem to be important for the function of the core region (P. Prehm and K. Jann, manuscript in preparation). The core may be substituted, e.g., on the KDO moiety with galactose in R2 (84) and with rhamnose in some K-12 strains (167) or on the terminal glucose of some K-12 strains with glucosamine or with N-acetylmannosaminuronic acid (225, 225a). The significance of these substitutions is not known. It is important to note that all core oligosac-


VOL. 41, 1977

685

TABLE 8. Structures of some neutral 0 antigens of E. colia 0 antigen 08

09 020 075

3 3

Man

Repeating unit 1,2 1 Man- Man-n

1,2 a

a

3

230

a

1 1,3 1,2 1,2 1,2 Man-e Man- Man- Man- Mana

4

Reference

Galp

a

1,2 a

a

1

Ribf

a

a

222 298

GlcNac1a Gal-1a

L-Rha-

40

1,4 Man

086

Gal

GalNac

0111

GalNac

270

I !Glc

T L-Fuc GlcNac-0

Glc-4 a

a 11,6

1

Gala

37

Col Col 6 Col (colitose), 3,6-dideoxy-L-galactose. If not otherwise indicated, the symbols for sugars used in this and the following tables are those suggested by IUPAC-IUB: Arch. Biochem. Biophys. 115:1-12, 1966.

charides are substituted with phosphate, phosphorylethanolamine, and pyrrophosphorylethanolamine. This together with the carboxyl groups of KDO renders the core region highly charged, may be of importance for the formation of superstructures on the cell wall and the maintenance of the outer membrane, and plays a role in the interaction of bacterial cells with bacteriophages or antibiotics. The full substitution is shown in Fig. 4 with the LPS ofE. coli B. R Antigens with an Incomplete Core E. coli B is an old laboratory strain with an rfa (incomplete) core. Treatment of S and R forms with mutagens often leads to R(rfa) mutants with incomplete core oligosaccharides. These incomplete structures are a great help in the elucidation of complete core oligosaccharides. In fact, the structures of R1, the K-12 core, and the B core were derived with the aid of such substructures (223, 225; Jann, manuscript in preparation). The structure of the core from E. coli B with all of its substituents is shown in Fig. 4. GENETICS OF 0 AND R ANTIGENS For a better understanding of the genetic determination of 0 antigens (LPS), the chemical structure of these molecules should be known. Information on this is given in the

chemical section of this article. The basic structural feature of LPS is shown in Fig. 2. The genetics of the Salmonella LPS have been studied extensively, and much knowledge has been gained about the biosynthesis of the core region and the O-specific polysaccharide by mating and transduction experiments, whereas less is known about E. coli in this respect. The Salmonella strain particularly studied is S. typhimurium. These studies, reviewed by Stocker and Makela (279), have shown that most of the genes determining the biosynthesis of the core region are present in a gene cluster (rfa) located in the cysE pyrE region on the Salmonella chromosome. The genes determining the O-specific repeating units are closely linked to the his operon and termed rfb. Mutations in either rfa or rfb result in a rough (R) phenotype. R mutants defective only in the rfb locus have a complete core, and R mutants defective in the rfa locus have a more or less incomplete core. The rfc gene between trp and gal is involved in the polymerization of the 0specific repeating units in some Salmonella 0 groups; if the rfc region is defective, only single repeating units are added to the core, resulting in an SR mutant instead of the normal smooth (S) form. Finally, a gene cluster termed rfe, close to ilv, appears to be involved in the synthesis of certain types of 0 antigens (as well as

686

ORSKOV ET AL.

BACTERIOL. REV.-

TABLE 9. Sugar composition of some acidic 0-specific polysaccharides from E. coli

E coli Q antigen

>

*0 d :

o4

+

58

U

p4 0 Z -~~ 24 41 54 56 58 59 69 80 83 87 96

-~

0

0 0

00

+

+ +

+

+

+

+

+ +

+

+

+ +

+

+

+ + +

+ +

+ + +

+

+

+ +

+ +

+

+ +

+

+

+ +

+

+ + +

+ + +

+

+

of enterobacterial common antigen [160, 161]); rfe- Salmonella mutants are phenotypically like rfb- mutants as no 0 specific side chain is formed. rtb, rfc, and rfe genes are thus participating in the synthesis of the 0-specific side chains. The genes concerned in LPS biosynthesis of E. coli, as far as studied, have been shown to relate to those in Salmonella. In 1962, it was shown by (rskov and Orskov (188) that the genes controlling O-antigen specificity in a number of smooth E. coli cultures were closely linked to the histidine locus. Strains of 0 antigens 06, 09, 025, 026, and 0100 were studied. Schmidt et al. (247) have similarly found that the genes for O-antigen synthesis in E. coli 08:K27 and 08:K42 are near his. This O-antigen-determining locus in E. coli is probably analogous to the rib locus in Salmonella. By mating some R strains isolated from E.

(

+

+ + (+

+

+

+ + + + +

+ + + +

+

+

+

+

+

+

+

+

+

+

+

(32)a a See footnote to Table 10.

+ + +

+

+

+ +

+ +

+

+

79 141

+

+

+

+

+

+

+

+

116 120 124 134 139 140

X

+

+

142ac

100

0

+

104 105 132ab

144 143

0

+ (+

+ +

+ + +

coli 08:K27 with S forms of the same strain as donors, it was demonstrated that the chromosomal site of the S -> R mutation in these strains was located close to mtl, a position similar to the rfa locus in Salmonella (248). In later experiments Schmidt (245) confirmed the presence of such rfa genes linked to mtl in the E. coli K-12 strain. One of the several described rfa genes in Salmonella is termed rfaL (279). It determines a component involved in the translocation of newly synthesized O-specific polysaccharide to a complete core. A gene equivalent to this rfaL gene was demonstrated in E. coli 08:K27 (250). The presence of a gene close to ilv equivalent to rfe in Salmonella has also been demonstrated in E. coli (251). By the transfer of an rfe defect from S. Montevideo to E. coli 08, 09, and 0100, recombinants were obtained that were blocked in the synthesis of O-specific polysaccharide (MAkela and Jann, unpublished data)


VOL. 41, 1977

687

TABLz 10. Structures of some acidic 0 antigens of E. coli 0 antigen

Repeating unit

(032)6

4

1,3

GkUA

FucNAc

1,3

GlcNAc

p

1,6

References

108, 286

1

Gal -2

1,4 Gk

068

1,4

3

GIcNAc

Dmitriev et al., manuscript in preparation

1 Man 2(3) Ac

1,4

Man

f13 RhaLA 1 1,4 1,2 1,4 3 -+Gal -> GlcNAc -> Rha -A Nba -, P-Glyc P-Glyc a

0100

0124

3 -

1,3

GalNAcp -

Gap

1,6

107; Prehm and Jann, manuscript in prep~~~~~~~~~~~aration 28

1 Gal 4

al1 1,4a GIcLA

1,6

Gk

GlcUA 0141

2 4

-

1,2 1,3 1,3 GlcNAc Man Man 1,6 j

GlcUA

1 T

1,3

1 1,3 Man - M GkNAc11

113

Rha Rha An 0 antigen related to 032 was described in the former test strain of antigen K87 (strain designation 145 = (032):K87:H45) (211): RhaLA (rhamnolactylic acid), 2-0-(1'-carboxyethyl)-L-rhamnose; GicLA (glucolactylic acid), 4-0-(1'-

carboxyethyl)-D-glucose.

TABLE 11. Sugar composition and occurrence of various R core types (244-246) Core des-

ignation GIcN R1 R2 R3 R4 K-12

1 1

Sugar composition As occ g GIc Gal Hepe m 3 S. sonnei, S. 3 2

46

24

3

GlcUA -

Reference

1

-Fuc

116

1,3 Gal K29

2

- man

1,3 1,3 1,3 1 > Gic > GlcUA Gal a ji B a

51

a1,4

a

Man 3

K30

M

1,2 1,2

Gic D

#

1,3

6

Pyruvate

1

105

GIcUA - > Gal -0

Gk1,3>GU 1,4)

1,2

K31K31

Gal -l Gic -> GlcUA -< Rha h -> h -> Gal Rha

K42

-

3

Gal

1,3 1,2 > GaIUA -

>

1 Jjann, unpubdata ->flished

1

-Fuc

115

GlcUA 2 1,2 1,3 -> GlcUA Man 4 1,6 T Rha

K85

1,3

Man

1,13

GlcNAc -- Man t

Man

1,3

-

1

113

GlcNAc 11

Rha

4 1 1,6 -f GlcUA -1,3 FucNAc -1,3 GlcNAc -> Gal

K87

286

1,4

X

Gic TABLz 15. Composition of some E. coli K antigens in 0 groups other than 08, 09, 020, and 0101 Neutral sugars K antigen

K1 K7

. Acidic sugar

Amino sugar

NANAa

K4

Mannosaminuronic acid GalUA

K54

GlcUA

GaIN

Gal

Glc

Man

Fuc

Rha

_

References

-

_

-

+

_ -

-

_ -

Threonine _ -

-

+

-

-

-

-

Jann, unpub-

-

-

-

-

+

+

lished data Jann, unpub-

4,168 166

lished data a

N-Acetylneuraminic acid.

K1+/Kl- (194). The K1+ polysaccharide has approximately 80 to 90% of the neuraminic acid residues O-acetylated, whereas the K1- variant has little or no detectable O-acetyl groups (see above, Test K antigens). The K100 antigen of E. coli, originally designated Kf147, also belongs to those K antigens in the 0 groups that are not 08, 09, 020, or 0101. It was reported to contain ribitol phosphate (233). Other sugar constituents, e.g., ribose, glucose, mannose, N-acetylglucosamme, N-acetylgalactosamine, and N-acetylmannosa-

mine, were also found in different preparations of the K1OO antigen isolated from various E. coli strains (233). The structure of the K100 antigen is not known. M Antigen As described above, the M antigen is not a type-specific K antigen, but, since it surrounds the bacterial cell as a thick slime layer, it fimctions as a capsular antigen. Because of its ubiquitous occurrence and the interesting regulation of its synthesis (123, 163, 164), it was stud-

VOL. 41, 1977

O AND K ANTIGENS OF E. COLU D-G.

0 / C

691

CH$

D-GA 3

D-GeI

311

_- -FuO -

4

1

3

L-FuC _ D-a

1

2-or 3 IAc x

FIG. 5. Structure of colominic acid (Ki antigen). = 8 (average) (from McGuire and Binkley [168D).

ied extensively (63, 64, 148, 282). The structure of the M antigen is shown in Fig. 6. What is commonly termed M antigen actually represents a group of acidic polysaccharide antigens. They all have the same polysaccharide backbone consisting of a hexasaccharide repeating unit. This may be substituted by formaldehyde (methylidene substitution), acetaldehyde (ethylidene substitution), or pyruvate (carboxyethylidene substitution). For all of these substituents, the terminal galactose of the side chain functions as the acceptor (64).

FIG. 6. Structure of the M antigen. R = CH3, isopropylidene substitution of galactose; R = H, ethylidene substitution ofgalactose (from Garegg et al. [631).

tR~ ~ ~,rfc

t~~~~ttiX~~~~~~a 75"'*, r

,

25

GENETICS OF K ANTIGENS The main gene loci on the coli chromosome determine two distinct groups of K antigens. One of these is linked to the his operon, and the 50 a other is linked to serA. The linkage map of the 7. Linkage map ofE. coli based on the E. coli FIG. E. coli chromosome is shown in Fig. 7. K-12 map of Bachmann et al. (2). Genes named outside the circle affect polysaccharide structures. K Antigens Determined near his Positions not accurately known are indicated by arrf, LPS biosynthesis; rfa, core; rfb, 0 repeating These K antigens have only been found in 0 rows. unit; rfc, polymerization of 0 repeating units; and groups 08, 09, and 020, and until now genes rfe, undefined. kps, polysaccharide K-antigen biohave been mapped for the following: K26 (188), synthesis; non, block in capsule formation (M antiK8, K9, K17, and K57 (Orskov and lrskov, gen); Ion (previously capR), regulation of capsular manuscript in preparation), and K27 (251a). polysaccharide synthesis (M antigen); capS, regulaThe gene locus controlling the synthesis of tor gene for capsular polysaccharide synthesis (M these antigens is closely linked to the rib gene antigen); thr, threonine; pro, proline; lac, lactose; galactose; trp, tryptophan; his, histidine; nal, cluster that determines the synthesis of the gal, nalidixic acid; srl, sorbitol; ser, serine; met, methio0-specific polysaccharides. nine; mal, maltose; str, streptomycin; xyl, xylose; Results obtained recently in crosses involv- mtl, mannitol; pyr, pyrimidine; ilv, isoleucine-vaing his mutants of 08:K8, 08:K25, 09:K9, and line; rha, rhamnose. 09:K57 as recipients have shown genetic linkSome of the K antigens of this group require age between genes for 0 and K antigens in these strains. The introduction of donor his an additional trp-linked gene locus for complete markers into these strains results in the acqui- expression (251a). This locus may be comparasition of donor O antigen (01 or 025) simulta- ble to the rfc locus in Salmonella that is necesneously with a loss not only of recipient 0 sary for the polymerization of 0-specific oligoantigens (08 or 09), but also of recipient K saccharide repeating units, and its product may antigens (K8, K9, K25, or K57) (0rskov and thus be a K-specific polymerase. These K antigens determined by a genetic Orskov, manuscript in preparation).

692

ORSKOV ET AL.

locus close to his were originally described as either L (K8 and K9), B (K25 and K57), or A (K26, K27, K29, K30, K31, and K42). K Antigens Determined near serA These antigens, of which K1, K4, K10, and K54 have been mapped (199, 208), are probably found in all 0 groups with the exception of 08, 09, 020, and 0101. The gene controlling their synthesis was termed kpsA, k for K antigen, ps for polysaccharide, and A because it is the first locus described that is involved in the structure of this type of K antigen. It was found recently (Schmidt, personal communication) that the synthesis of the K7 antigen is controlled by a gene locus that is not linked to the his operon. However, it was not established whether the K7 antigen is controlled by the kpsA gene. From a cross between a K10- donor and a K54- recipient, some recombinants were obtained that expressed the latent K antigen of the donor. These results indicated that at least two genes were operating to give a K+ phenotype, one concerned with the structure (kpsA) and the other concerned with the expression of the K antigen (188, 199). Genetic manipulation allows the construction of E. coli strains with both types of K antigens present on one cell. Thus, in crosses between serA mutants of E. coli 08:K8 or 09:K57 strains as recipients and E. coli 025:K10 as the donor, the recombinants expressed both donor and recipient K antigens. Because of the stable presence of two K antigens, the finding was explained as an indication of the fact that the K8 and K57 antigens map at a locus other than K10 (208). Form variation has been described in the K1 antigen (see above). M Antigen Although we do not consider the M antigen as belonging to the true K antigens because it is found with the same specificity in many Enterobacteriaceae strains, the regulator genes that control the synthesis of this antigen (163, 164) should be mentioned. Mutations in two regulator genes designated lon and capS lead to a derepressed synthesis of several enzymes involved in the synthesis of the M antigen and thereby to an overproduction of M-antigen polysaccharide. The structurel genes for some ofthe enzymes are mapped (152). Two of the enzymes also participate in reactions not necessarily connected with the synthesis of the M-antigen polysaccharide, and thus other structural genes that are more exclusively associated with Mantigen synthesis may be found in an M-anti-

BACTERIOL. REV.

gen polysaccharide operon (165). Another mutation located near his, termed non-9, inhibits capsule formation (M antigen) (226). K Antigens Determined by Plasmids The fimbria-like K antigens K88 and K99 are proteins (105a, 278). The determinants of both antigens are transferable and located on plasmids (202, 211, 263).

BIOSYNTHESIS OF 0 AND K ANTIGENS The principles of the biosynthesis of microbial polysaccharides were established in a number of Salmonella strains (172, 175, 215, 217, 235). It was found that oligosaccharide repeating units are first assembled on a carrier lipid in the cytoplasmic membrane. They are then polymerized to the polysaccharide chain that is still attached to the carrier lipid. In the biosynthesis of LPS, the polysaccharide is transferred from the carrier lipid to the core lipid A part of the molecule. The reaction(s) in which free polysaccharides are detached from the carrier lipid is not known. The carrier lipid is a polyprenolphosphate (99, 310), which, after liberation during the polymerization and release of the polysaccharide, is recycled in the synthetic process. The intermediary participation of polyprenolphosphates (bactoprenolphosphate and dolicholphosphate) was also reported for the synthesis of the cell wall glycopeptide (25, 66, 98), capsular polysaccharides of Klebsiella (178, 283, 292, 293), and yeast and bacterial mannans (151, 242) as well as polysaccharides of plants (94, 299) and higher organisms (5, 6, 220, 221). It was therefore thought that all polysaccharides would be synthesized by the same mechanism in which membrane-bound lipids function as carrier molecules. Relatively little is known about the biosynthesis of polysaccharide antigens in E. coli. The K30 antigen of E. coli 09:K30 was shown to be synthesized via lipid-linked oligosaccharides by a mechanism similar to that described above (D. K. Chandler and K. Jann, manuscript in preparation). It is likely that other K antigens, which are related to the K30 antigen (Tables 13 and 14), are synthesized in the same way. The in vitro synthesis of the 0-specific polysaccharide of E. coli 0111 was studied by Edstrom and Heath (38). They found that the sugar constituents are incorporated into cell envelope fractions in the sequence galactoseglucose-glucosamine-colitose. No oligosaccharides were detected in the synthesis of the 0111 polysaccharide, and no lipid-bound intermediates could be found.

More recently the biosynthesis of the E. coli

VOL. 41, 1977

09 polysaccharide was studied with membrane vesicles (membrane mixtures and isolated cytoplasmic membrane [see reference 216]). The 09 polysaccharide, which is a mannan, is synthesized from guanosine 5'-diphosphate-mannose without the participation of a lipid intermediate. The synthesis is independent of bivalent metal ions, in contrast to the situation for Salmonella andE. coli 0111, where the incorporation of galactose (the first sugar in the sequence) depended on the presence of magnesium ions. Although extraction with organic solvents failed to show an intermediate of the synthesis, such an intermediate could be detected by polyacrylamide gel electrophoresis in the presence of dodecyl sulfate, as was also shown independently in the synthesis of teichoic acid (55, 56). The synthesizing system can be extracted with 0.01% Triton X-100, which is an advantage for the further study of the system (Winter and Jann, manuscript in preparation). The 09 polysaccharide thus seems to be synthesized in a single chain mechanism that is different from the one described above. The mechanism of 0-polysaccharide biosynthesis found in E. coli 09 is also operative in E. coli 08 and Klebsiella strains 03 and 05 (142; H. C. Flemming and K. Jann, manuscript in preparation). In this context it is interesting to note that in E. coli strains 08 and 09 no rfc locus (determining the polymerase for lipidlinked oligosaccharides [see reference 279]) was found.

MULTIPLE OCCURRENCE OF POLYSACCHARIDE ANTIGENS E. coli strain D227, 08:K87 was originally believed to contain a neutral cell wall LPS and, in addition, an acidic capsular polysaccharide (108, 113, 210, 286). It was found later (Jann and Jann, manuscript in preparation) that the K87 antigen was actually an acidic LPS, so the strain contains two LPS in the outer membrane. As described above, the acidic LPS of this strain is still considered to be a K antigen since no K87 antibodies are demonstrable in an 0 antiserum. The presence of two LPS, one neutral and one acidic, is also found in strains (isolated from calves) ,erotyped as 08:K85. The K85 and K87 specificities also occur in E. coli 0141:K85 (strain RVC2907) and (032):K87.(strain 145), respectively (Jann, unpublished data). This corresponds to the fact that extracts from these strains gave only one anodic line in immunoelectrophpresis (194). The structures of K85 and K87 Are shown in Table 14. One can speculate whether, during evolutionary processes involving host-parasite interactions in terms of invasion and defense, anti-


693

genic drifts occur, such as an R strain being converted to E. coli 08 which has acquired a neutral O-specific polysaccharide, further to E. coli 08:K87 with an additional acidic (lipo)polysaccharide, and finally to (032):K87 which seems to have lost the neutral (lipo)polysaccharide. A similar sequence can be envisaged with the strains R-08-08:K850141:K85. In fact, it is possible that all E. coli strains with O-specific acidic LPS (Tables 9 and 10) may have arisen in such a fashion, finally arriving at a stage where the properties of a cell wall antigen (lipid A), important for the maintenance of the outer membrane, and those of a capsule (acidic polysaccharide), important for counteracting phagocytosis and adverse influences on the ecosystem, could be combined into one molecule.

PATTERNS OF POLYSACCHARIDE ANTIGENS The results of the chemical and immunochemical studies are compiled in Table 16. The 0 antigens that are shown in the first four columns are neutral LPS. Their O-specific polysaccharide chains are mostly heteropolysaccharides; only the 0 antigens 08 and 09 have O-specific homopolysaccharides. In immunoelectrophoresis in agar, all ofthese neutral LPS show 0 lines, which, due to endosmosis, are located toward the cathode. Capsular antigens may occur together with these LPS (as in the first three columns of Table 16). These are either acidic polysaccharides (columns 1 and 2) or acidic LPS (column 3). In immunoelectrophoresis they all give rise to an additional K line, when anti-OK sera are used. K-antigenic acidic polysaccharides of relatively high and relatively low molecular weights can be differentiated in immunoelectrophoresis by their different electrophoretic mobilities. It is, however, not possible to differentiate with this technique acidic polysaccharides of high molecular weight from acidic LPS. Although the acidic LPS have polysaccharide chains of the usual small size (about 10,000), they acquire a very high apparent molecular weight by micelle formation via their lipid A moiety. The 0 antigens that are shown in columns 5 and 6 of Table 16 are acidic LPS. In some cases (column 6) an additional capsular polysaccharide, probably of low molecular weight, is found. On the whole, the results of chemical and immunoelectrophoretic studies are in good agreement. The designation of the various immunoelectrophoretic groups, as introduced in Fig. 1, are therefore also given in Table 16. Some antigenic patterns shown in Table 16 are also found in other genera. Thus, E. coli

694

ORSKOV ET AL.

~v

BACTZRIOL. Rzv.

TABLE 16. Polysaccharide patterns (combinations of polysaccharide antigens) in E. coli based on chemical and immunoelectrophoretic characterizations O In 0 groups In 0 groups In many 0 In many 0 In many 0 groups groups 09, 08, 09, groups 08, 0101, (020) 0101, (020) Neutral LPS Neutral LPS Neutral LPS Neutral LPS Acidic LPS Acidic LPS Acidic LPS Acidic capsuAcidic capsular Acidic capsular polysacpolysacchalar polysaccharide charide ride

Chemical char- In many groups acterization

Immunoelectro-

phoretic characterization

0 line: cathodic slow K line: anodic fast

Immunoelectro- 1A phoretic _ groups

0

0 line: cathodic slow K line: anodic slow

As described in the legend to Fig. 1.

CHEMICAL BASIS OF ANTIGENIC SPECIFICITY A number of reviews have been written on the immunochemical aspects of carbohydrate antigens (112, 155-157). Therefore, only information that is essential for the understanding of specificities of E. coli polysaccharide antigens and their cross-reactions with other polysaccharides will be given here. Extensive studies in many laboratories revealed that the antigenic specificities of polysaccharides in most cases center around single hexose units and extend along the polysaccharide chain over regions of different lengths. The sugar unit that contributes most to the antigenic specificity (contributes the highest increment of binding energy between the antigenic determinant and the binding site of a homologous antibody) was termed immunodominant sugar (156). In structural formulas this is usually indicated by an arc, according to a suggestion by A. M. Staub (see references 112 and 156). Immunodominant sugars may be branch substituents of a polysaccharide main chain, such as colitose in the 0111 antigen (Fig. 8) and galactose in the K27 antigen (Fig. 8). They may also be part ofthe polysaccharide chain, as with the K30 and K42 antigens (Fig. 8). In polysac-

2b

lBb

lBa

strains with an antigenic pattern (combination of polysaccharide antigens) shown in column 2 closely resemble Klebsiella, and those with an antigenic pattern shown in column 5 resemble certain Shigella strains (S. dysenteriae and S. boydii). In fact, identity of surface antigens and serological specificity was found between some E. coli and Shigella strains (27, 28, 141, 275) as well as between some E. coli and Klebsiella strains (21, 153, 222, 230).

line: ca- 0 line: anodic 0 line: anodic slow thodic slow slow K line: anodic fast

f

a

2a

v

-GlcNac z

1.60

Gc

a

age

b

_

c

13 2 A --. Gal GlcUA 2.

d

3

Glc

Gal

GlcUA

N

_ Gal -.GlcNac 2 Gic =* Go la

rIa

ag( I

GlcUA 2 Fuc

v Fuc -_-* Gic

1.3 Gal 1.2 GlcUA -. 13M Gal -.Mon-Man to -3

. GOIUA -- Fuc --

Gal

-

GoUA --Fuc--

FIG. 8. Antigenic carbohydrate determinants of the 0111 antigen (a), K27 antigen (b), K30 antigen (c), and K42 antigen (d). Immunodominant sugars are indicated by arcs, and repeating oligosaccharide units are indicated by brackets.

charide antigens these determinants repeat along the polysaccharide chain. Also, several distinct antigenic determinants may alternate in a polysaccharide. In addition to structural determinants, which are expressed by nature, sequence, and linkage of a few sugar residues, there may exist another type of determinant that could be termed "conformational determinant" (258). It depends on structures of secondary order, i.e., specific spatial arrangements of a polysaccharide such as helix or coil formation. The term conformational determinant was used by Girard and Staub (68) in the immunochemical analysis of the LPS from Salmonella johannesburg. The antigenic specificities of the 08 and 09 antigens of E. coli seem to be based on a special conformation of the O-specific polysaccharides (a-mannans). In spite of the similar

VOL. 41, 1977

structures of the 08 and 09 antigens, which contain identical sequences, there is no crossreaction at all between these two coli antigens. Both (and also the 03 and 05 antigens ofKlebsiella that cross-react with E. coli 09 and 08, respectively) lose their 0 specificity completely when they are degraded with mild acid, a procedure that liberates the carbohydrate from the (micelle forming) lipid A (230; Jann, unpublished data). It was found (Jann and Jann, unpublished data) that the substitution of the mannns with stearoyl groups (1 to 2 residues per polysaccharide molecule of about 10,000 daltons) restored the serological activity. Since stearoyl groups can associate with each other (micelle formation) or with hydrophobic regions of the polysaccharide (alteration of conformation), this result can be taken as an indication of a secondary structure in the 0-specific mannans. In accordance with this, no oligosaccharides with 08 or 09 specificity were found. Antigenic specificity may also be due, at least in part, to noncarbohydrate substituents such as pyruvic acid in the K29 antigen (Table 14) and the M antigen (Fig. 6) or glycerophosphate in the 0100 antigen (Table 10).


695

are different, even though infantile and piglet diarrheas have many other common traits. E. coli from Humans In Table 18 the findings from extra-intestinal infections are summarized. Only few of the many reports published are included, but those listed should be representative. Only the most frequent 0 groups have been included. The percentages for the single 0 groups differ, but it is probably true that with the use of those 0 sera that correspond to the 0 groups listed, it should be possible to determine the 0 antigen of at least half of all coli strains from urinary and other extra-intestinal infections. It should be noted that these frequent 0 groups are the same as those found to be prevalent in the normal intestine, but the percentages for feces are lower. Similarly, only a few polysaccharide K antigens (K1, -2, -3, -5, -12, and -13) are found with high frequency in urinary tract infections and other extra-intestinal infections (121b, 159, 296). Possibly, invasiveness is paralleled by a further selection of strains already frequent in the intestine. Today most authors agree that the same 0 groups are frequent in the normal healthy intestines and extra-intestinal infections. However, in the special discussior about the involvement ofE. coli in urinary ract infections, two seemingly different views have been expressed. One is named the "special pathogenicity theory," stating that some serovars are more frequent in the urinary tract than in the intestinal tract due to the special pathogenicity of such strains. The other, called the "prevalence theory," says that the serovars found in urinary infections represent the most prevalent types in the gut (81). Unfortunately, comparable data that could make possible a definite choice between the two theories are missing. Some authors have indicated that differences in O-group distribution exist, e.g., between Europe and the United States (71) and even within different areas in London (80). Others have suggested that prevalence rates of certain 0 groups could vary with time in the same area (159). Even though such differences do occur, they can probably be explained by the use of different techniques and sera. The overall picture of a similar E. coli group prevalence in Western developed countries in temperate climates (301) is not disturbed by these discrepan-

CHEMICAL BASIS OF CROSS-REACTIONS The molecular basis of some cross reactions in which polysaccharide (O and K) antigens of E. coli take part has been elucidated (Table 17). OF SEROLOGICAL PREVALENCE GROUPS AND TYPES OF E. COLI IN PATHOLOGICAL CONDITIONS IN HUMANS AND ANIMALS Little is known about the normal distribution of different serogroups. Best examined are strains from humans and common domestic animals in temperate climates, whereas we know little about the distribution of serogroups and types in warmer geographical regions. Most examinations have been carried out in developed countries. It is not yet possible to state whether there are significant differences in 0-group prevalence between different animal groups. Recently it has been stated that this is not likely between humans and cattle (88). We agree that many frequent 0 groups are the same, but we would hesitate, on the other hand, to draw this conclusion until further examinations have been carried out. It is astonishing, for example, that 04 and 06 strains hardly occur among frequent cattle strains. When the prevalent 0 cies. The results of serotyping carried out on difgroups from diseased animals and humans are compared, it is evident that most of the strains ferent types of diarheal diseases in humans are associated with diarrhea in pigs and humans summarized in Table 19. Part a lists the most

TABLE 17. Cross-reactions of some E. coli polysaccharide antigens E. coli antigen

Probable common antigenic determinant

Cross-reacting antigen

08a

Klebsiella 05"

O9a

Klebsiella 03c

A Man and/or conformational A a Man and/or a

019ab

Klebsiella Old

3

020a

Klebsiella 04e

058' 01249

S. dysenteriae 3 S. dysenteriae 5

KI'

N. meningitidis b

Terminal 3-O-(1'-carboxyethyl)-L-Rhaf Terminal 4-0-(1'-carboxyethyl)-D-Glc# 8 -* NANA and/or conformational

K4

Pneumococcus XII

GlcUA possibly ManNacUA

K7

Pneumococcus III

X Glc possibly ManNacUA

K8

Pneumococcus VIII Pneumococcus XIX Pneumococcus XXII Pneumococcus XXII

4A, Glc

K26

Pneumococcus II

a

*

2*

a

Man and/or

Gal

-

a

Man and/or conformational

a

Rib and/or Xk Gal X

a

Glc A. Glc Glc and/or Gal

GlcUA (or terminal in K26) and/or A Rha

Pneumococcus XXIII

A Rha - GlcUA (or terminal in K30) A Rha (or terminal in K31)

K42k

Pneumococcus XXV

A GalUA and/or 3 Gal

K54

Pneumococcus III Pneumococcus VI Pneumococcus II Pneumococcus X

A GlcUA -+ Gal (2 position in VI or terminal in K54) Terminal GlcUA 2 GlcUA (in V) A GlcNac

Pneumococcus VIII

Gal < GlcUA

K30k K31

K85k

Pneumococcus VI Pneumococcus II

Pneumococcus V

K87

Structures are given in Table 8. b Structures are given in reference 153. Structure is given in reference 21. d Structure is given in reference 9. e Structure is given in reference 10. ' Dmitriev et al., manuscript in preparation. Structure is given in Table 10. h Structure is given in references 27, 28, and 141. 1 Structure is given in Fig. 5. J The K7 antigen contains glucose and N-acetylmannosaminuronic acid, the pneumococcal polysaccharides III and VIII contain glucose and glucuronic acid. The charge of the acidic sugar may also play a role in cross-reactions. k The data for K30, K42, and K85 polysaccharides are from reference 91, and the results for the other K antigens are unpublished results of Heidelberger and Jann. For the composition or structures of the pneumococcal polysaccharides, see reference 90. a

c

696

VOL. 41, 1977

0

common enteropathogenic E. coli serovars. These types can be associated with infantile diarrhea. They were all originally isolated from severe outbreaks of infantile diarrhea in institutions and are also today most frequently found in such places. Only 0 groups have been listed, i.e., 026 and 055, since complete serotyping of such strains has been carried out to a very limited extent. However, it was found (45, 185) that some O:H combinations (055:H6, 0111:H2, 086:H34) are much more frequently associated with clear-cut outbreaks than oth18. E. coli 0 groups from extra-intestinal infections in humans 0 groups' Infection References Urinaryb 01, 02, 04, 06, 7, 43, 62, 81, 159, 07, 08, 09, 177, 227, 294, 011, 022, 025, 296

TABLz

Septicemia

Other Neonatal men-

ingitisc

Feces (healthy and adults children)

062, 075 01, 02, 04, 06, 07, 08, 09, 011, 018, 022, 025, 075 01, 02, 04, 06, 08, 09, 011, 021, 062 01, 06, 07, 016, 018, 083 01, 02, 04, 06, 07, 08, 018, 025, 045, 075, 081

43, 74, 191

43,79,295,296 231, 241 128, 227, 255, 295, 296, 301, 305

Since the prevalence rates have been compiled from many different investigations, the 0 groups are listed numerically. ' A limited number of polysaccharide K antigens are found to dominate among strains from urinary tract infections: KI, -2, -3, -5, -12, and -13. The same K antigens are also frequently found among stool isolates from healthy persons (121b, 159, 296). e The prevalent 0 groups listed here are characterized by having the same K antigen, K1, when found in this disease.


697

ers. The LPS of these strains are neutral, and they have no acidic polysaccharide K antigens (190), which seems to be in accordance with their noninvasive character. These serovars can also be found in institutionalized infants with no diarrhea, and the cause of pathogenicity in these strains cannot yet be explained. Fimbria-like antigens such as those found in piglet and calf strains (211, 212, 263) have not yet been described; neither has a definite proof for a general production of enterotoxin in such strains been given. Recently there has been a tendency to reject the importance of these special serovars completely because few such types examined have been enterotoxigenic. It should be remembered, however, that E. coli serotyping has been of great value for the elucidation of many serious outbreaks and epidemics of infantile diarrhea. On the other hand, it is of course true that serotyping of these so-called enteropathogenic types probably is of limited value for the examination of sporadic cases of infantile diarrhea. In part b of Table 19, 0 groups of the dominant E. coli flora of sporadic cases of diarrhea in infants (not colonized by the so-called enteropathogenic serovars) are listed. It is apparent that the common fecal 0 groups are found (186), a result that might indicate that a possible etiological role of these strains is not linked to

the 0 group. Part c of Table 19 lists some O:H serovars frequently found among enterotoxigenic strains isolated from diarrheal diseases in adults and children, mostly in warm climates. Some of these O:H types have been isolated from the socalled traveler's diarrhea (196). Recently it has been shown that enterotoxigenic strains may also be the cause of food- and water-borne dis-

TABLz 19. E. coli 0 groups and O:H types from intestinal infections in humans Infection

Infantile diarrhea (a) From outbreaks, mostly institutions in developed countries (b) Sporadic cases excluding enteropathogenic types Diarrhea in adults and children (c) Enterotoxigenic strains, mostly sporadic cases

(d) From food-borne disease

0 groups and O:H serotypes

References

020, 026, 044, 055, 086, 0111, 0114, 0119, 0125, 0126, 0127, 0128, 0142, 0158, 0159 01, 02, 04, 06, 08, 015, 021, 051, 075, 085

39, 45, 79a, 134, 185, 237a, 238, 256, 288a

06:H16, 08:H9, 015:H11, 025:H42, 078:H11, 078:H12, 0128:H7, 020:H-

i96, 238, 205a, manuscript in

78, 186

preparation

06, 08, 015, 078, 0124, 162a, 239 0149 (e) Dysentery-like disease 028ac, 0112, 0124, 0136, 179, 240 0143, 0144 References given are mostly to reviews in which further references can be found.

BACTZRIOL. REV.

0RSKOV ET AL.

698

eases (239). A more detailed typing, also including K-antigen and biotype determinations, shows that special serofermentative O:K:H types are frequently associated with enterotoxigenicity (205a). Serogroups associated with dysentery-like diseases are listed in part e of Table 19. In geographical regions where S. dysenteriae is a common type of diarrhea, these E. coli serovars are also found to be associated with dysentery. By the usual laboratory procedures, they will be labeled E. coli, but most of them have close O-antigen relationships to different Shigells serogroups (39). They have the same invasiveness as Shigella strains, are positive in the Sereny test (261), and cause primarily a disease of epithelial layers of the colon. The last column of Table 19 records a number of references to papers where O-group distribution in normal healthy persons has been examined. It is seen that some 0 groups can be called the common 0 groups, among which those listed in parts a and e of Table 19 are not found. The so-called enteropathogenic serotypes from infantile diarrhea and the Shigella-like serovars are thus rare in healthy intestines. The 0 groups from enterotoxin-determined diarrhea are not rare, but there are indications that some of these strains represent well-defined combinations of special 0, H, and K antigens and have special fermentation patterns that are not common in healthy intestines (196, 205a).

E. coli from Cattle, Pig, and Poultry When Jensen in 1893 (117) described the etiology of diarrhea and septicemia in calves, it was the first attempt to associate E. coli bacteria with a special disease syndrome. Table 20 summarizes the knowledge about the association between E. coli 0 groups and TABLE 20. E. coli 0 groups from cattle Infection Diarrhea (white scours) and septicemia in calves Healthy feces

Mastitis

calf

0 groups 01, 02, 08, 09, 015, 020, 026, 055, 078, 087, 0101, 0114, 0115, 0117, 0137 01, 02, 04, 06, 08, 09, 011, 013, 015, 018, 020, 022, 023,

036, 0101, 0107, 0116, 0117, 0123, 0153 02, 08, 09, 021, 081, 086

References 12, 22, 53, 65, 70, 120, 154, 162, 228, 253, 264, 266,297,309 70a, 104a, 104b, 309

52

certain diseases in cattle. As in the previous similar tables in this paper, results from different investigations have been combined because a general prevalence of some frequent 0 groups is apparent. The differences in prevalence found among various reports may be real but can, most simply, be explained by differences in the selection of strains and in the typing techniques. It is apparent from Table 20 that few examinations of the E. coli 0 groups in healthy calves have been carried out. If the 0 groups from sick and healthy calves are compared, only few significant differences in 0-group prevalence can be found, the only important one being that Wramby (309) never found 078 strains in healthy animals. Fey (54) confirmed this rare occurrence of 078 strains in healthy

animals.

Table 21 shows the E. coli 0 groups that are frequently found in young pigs. For an extensive review, see Sojka (266). The strains from diarrhea, especially in newborn piglets, very often carry the fimbria-like antigen K88. Most of the 0 groups listed in Table 21 are usually found as well-defined serofermentative types, i.e., 0138:H14 (earlier 0138:K81:H14), with a typical fermentative pattern independent of their geographical place of isolation. It should be mentioned that 02:K1 in most reports is the dominant O:K group from septicemia, and next in frequency are 078 and 01:K1 (Table 22). For an extensive review, see Sojka (266).

SUMMARY AND CONCLUSION A thorough and precise knowledge about the surface structures ofE. coli is evidently important. The E. coli group (species) consists of a high number of more or less closely related serovars or serofermentative types. These can be differentiated in many ways and by many techniques, but until now the most practical and useful procedure has been based on different antigenic surface structures. The all-important foundation for this work was laid down by Kauffmann 30 years ago, and since then sero-

TABLE

21.

Prevalent E. coli 0 groups from pig diseases

0 groups

References

Infection Diarrhea (mostly enterotoxigenic strains from

83,136,143, 228, 08:K87, 045, 265, 268, 284, 0101, 0138, 308 0139, 0141, 0147, 0149, 0157

piglets) Edema disease

0138, 0139, 0141

136, 143, 268

0 AND K ANTIGENS OF E. COLU

VOL. 41, 1977

TABLz 22. E. coli 0 groups from poultry Infection Coli septicemia (septic pericarditis, air sacculitis, etc.) HjArre's disease, coli

granu-

loma disease, mucoid strains

0 groups 01, 02, 08, 071, 073, 078,088

References 86, 87, 92, 93, 267, 285

08, 09, 016

309

699

antigens in pathogenicity. An interesting phenomenon is the prevalence of certain E. coli strains in humans or animals and their correlation with pathogenicity. An idea evolving from comparative studies is that of specific recognition of and interaction between bacteria and mammalian cells. It is conceivable that histocompatibility antigens or surface antigens expressed on cells of certain tissues play a role in the attachment of bacteria as the first step in an infection. This may be due to (i) antigens common to the bacterial and mmmalian cell surface causing impaired recognition of the bacteria or to (ii) the presence on the mammalian cell of antigenic structures acting as receptors for complementary bacteria surface antigens. In the interaction of host and parasite, defense mechanisms developed by the host were counteracted by the bacteria, with the production of protective surface structures. As new surface antigens evolved, the production of previous ones, which may have proved to be insufficient, is probably discontinued. Thus, such an evolutionary adaptation to a changing ecosystem may have produced antigenic drifts, and this can b1e considered as one of the reasons for the enormous diversity in the patterns of E. coli antigens. Their study by chemical and genetic approaches, as well as the investigation of the biosynthetic pathways leading to the polysaccharide antigens in question, is certainly exciting. Furthermore, it can easily be appreciated that such studies are essential for our understanding not only of the pathogenicity of E. coli but also of bacterial pathogenicity at large.

typing has been carried out mainly based on the 0 antigens (0-specific polysaccharide of LPS) and the H antigens (protein of flagella). Although the K antigens have been known to exist during this same period, they have played a minor role in general serotyping. The introduction of some new techniques, mainly based on immunoprecipitation in gels, and the application of chemical and genetic methods have given us a more coherent picture of the capsular or microcapsular (acidic) polysaccharide K antigens. This development has made a more reliable serotyping of these antigens possible. Only few strains have been examined for their polysaccharide K antigens, and only future research will show how many more serologically different capsular antigens exist; the present number is about 70. So far, the protein surface structures of E. coli have played a small role in serotyping procedures, with the exception of the two important K antigens, K88 and K99. Recent interest in many laboratories in fimbrial and pili structures, because of their possible role as adhesive organs, may increase the number of such proACKNOWLEDGMENTS tein K antigens considerably. We thank Agnete Thorborg and Liselotte ter The importance of Escherichia strains, both Haak for their tireless efforts during the many as normal inhabitants of the intestinal tract of the production of the manuscript. and as intestinal and extra-intestinal patho- stages gens, has been recognized for many years. LITERATURE CITED However, only recently have we begun to un1. Anderson, E. S., and A. H. Rogers. 1963. Slime derstand the different mechanisms underlying polysaccharides of the Enterobacteriaceae. the diverse aspects of Escherichia pathogenicNature (London) 198:714-715. ity. The pathogenicity of E. coli is a very com2. Bachmann, B. J., K. B. Low, and A. L. Taylor. plex phenomenon, in which the chemical na1976. Recalibrated linkage map of Escheture of the bacterial cell surface plays a promirichia coli K-12. Bacteriol. Rev. 40:116-167. 3. Baron, L. S., P. Gemski, E. M. Johnson, and nent part. The negative surface charge due to S. Wohlheiter. 1968. Intergeneric bacterial acidic capsular or cell wall polysaccharides may matings. Bacteriol. Rev. 32:362-369. be important for the penetration of bacteria into 4. Barry, G. T., and W. F. Goebel. 1957. Colomthe tissues. The presence of fimbriae on bacteinic acid, a substance of bacterial origin reria enables them to attach to the mucoid epithelated to sialic acid. Nature (London) 179:206lial cell layers, such as intestinal linings. Spe208. cial immunological techniques and chemical 5. Behrens, N. H., A. J. Parodi, and L. F. Leloir. analyses have opened the door to an under1971. Glucose transfer from dolichol monophosphate glucose: the product primed with standing of the role played by bacterial surface

700

6.

7.

8.

9.

10.

11.

12.

13.

14.

15. 16.

17.

18.

19.

ORSKOV ET AL. endogenous microsomal acceptor. Proc. Natl. Acad. Sci. U.S.A. 68:2857-2860. Behrens, N. H., H. Carminiatti, R. J. Staneloni, L. F. Leloir, and A. I. Cantarella. 1973. Formation of lipid-bound oligosaccharides containing mannose. Their role in glycoprotein synthesis. Proc. Natl. Acad. Sci. U.S.A. 70:3390-3394. Bergstrom, T., K. Lincoln, F. Orskov, I. 0rskov, and J. Winberg. 1967. Studies on urinary tract infections in infancy and childhood. VIII. Re-infection vs relapse in recurrent urinary tract infections. J. Pediatr. 1:13-20. Bettelheim, K. A., F. M. Bushrod, M. E. Chandler, and R. E. Trotman. 1975. An automatic method for serotyping Escherichia coli. Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. Reihe A 230:443-451. Bjorndal, H., B. Lindberg, and W. Nimmich. 1971. Structural studies on Klebsiella 0 group 6 polysaccharide. Acta Chem. Scand. 25:750. Bjdrndal, H., B. Lindberg, I. Lonngren, K. Nilsson, and W. Nimmich. 1972. Structural studies on the Klebsiella 0 group 4 lipopolysaccharide. Acta Chem. Scand. 26:1269-1271. Boivin, A., and L. Mesrobeanu. 1935. Recherches sur les antigenes somatiques et sur les endotoxines des bact6ries. I. Considerations generales et expose des techniques utilisees. Rev. Immunol. 1:553-569. Bokhari, J. M. H., and F. 0rskov. 1952. 0 grouping of E. coli strains isolated from cases of white scours. Acta Pathol. Microbiol. Scand. 30:87-89. Bradshaw, M., R. Schneerson, J. C. Parke, Jr., and J. B. Robbins. 1971. Bacterial antigens cross-reactive with the capsular polysaccharide of Haemophilus influenzae type b. Lancet i:1095-1096. Braun, V. 1975. Covalent lipoprotein from the outer membrane of Escherichia coli. Biochim. Biophys. Acta 415:335-377. Braun, V., and K. Hantke. 1974. Biochemistry of bacterial cell envelope. Annu. Rev. Biochem. 43:89-121. Brinton, C. C. 1965. The structure, function, synthesis and genetic control of bacterial pili and a molecular model for DNA and RNA transport in gram negative bacteria. Trans. N.Y. Acad. Sci. Ser. II 27:1003-1054. Brinton, C. C., Jr., P. Gemski, and J. Carnahan. 1964. A new type of bacterial pilus genetically controlled by the fertility factor of E. coli K12 and its role in chromosome transfer. Proc. Natl. Acad. Sci. U.S.A. 52:776-783. Bussard, A., and J. Huer. 1959. Description d'une technique combinant simultanement l'electrophorese et la precipitation immunologique dans un gel: l'6lectrosyn6rese. Biochim. Biophys. Acta 34:258-260. Byelaya, J. A. 1965. Application of the immunoelectrophoretic method in agar for study-

BACTEXIOL. REV.

20.

21.

22.

23.

24. 25.

26.

27.

28.

29.

30. 31.

32.

33. 34.

35.

ing the somatic antigen of dysentery bacilli. Zh. Mikrobiol. Epidemiol. Immunobiol. No. 9, p. 55-61. Ceppellini, R., and M. Landy. 1963. Suppression of blood group agglutinability of human erythrocytes by certain bacterial polysaccharides. J. Exp. Med. 117:321-338. Curvall, M., B. Lindberg, J. Longren, and W. Nimmich. 1973. Structural studies on the Klebsiella 0 group 3 lipopolysaccharides. Acta Chem. Scand. 27:2645-2649. Dam, A. 1960. Serological 0 grouping of strains of E. coli isolated from cases of coli septicaemia in calves. Nord. Veterinaermed. 12:321342. Datta, N. 1975. The diversity of the plasmids that determine bacterial conjugation. Proc. Soc. Gen. Microbiol. 3:26-27. Davis, B. R., and W. H. Ewing. 1958. Six new Escherichia coli H antigens. Can. J. Microbiol. 4:517-519. Dietrich, C. F., A. V. Colucci, and D. L. Strominger. 1967. Biosynthesis of the peptide glycan of bacterial cell walls. V. Separation of protein and lipid components of the particular enzyme from M. lysodeikticus and purification of the endogenous acceptor. J. Biol. Chem. 242:3218-3225. Dmitriev, B. A., L. V. Backinowsky, V. L. Lvov, N. K. Kochetkov, and I. L. Hofman. 1973. Immunochemical studies on Shigella dysenteriae lipopolysaccharides. Eur. J. Biochem. 40:355-359. Dmitriev, B. A., L. V. Backinowsky, V. L. Lvov, N. K. Kochetkov, and I. L. Hofman. 1975. Somatic antigen of Shigella type 3. Structural features of specific polysaccharide chain. Eur. J. Biochem. 50:539-547. Dmitriev, B. A., V. L. Lvov, N. K. Kochetkov, B. Jann, and K. Jann. 1976. Cell-wall lipopolysaccharide of the "Shigella-like" Escherichia coli 0124. Structure of the polysaccharide chain. Eur. J. Biochem. 64:491-498. Drach, G. W., W. P. Reed, and R. C. Williams. 1971. Antigens common to human and bacterial cells: urinary tract pathogens. J. Lab. Clin. Med. 78:725-735. Duguid, J. P. 1959. Fimbriae and adhesive properties in Klebsiella strains. J. Gen. Microbiol. 21:271-286. Duguid, J. P. 1964. Functional anatomy of E. coli with special reference to enteropathogenic E. coli. Rev. Latinoam. Microbiol. 7:116. Duguid, J. P. 1968. The function of bacterial fimbriae. Arch. Immunol. Ther. Exp. (Engl. Transl.) 16:173-188. Duguid, J. P., and R. R. Gillies. 1957. Fimbriae and adhesive properties in dysentery bacilli. J. Pathol. Bacteriol. 74:397-411. Duguid, J. P., E. S. Anderson, and I. Campbell. 1966. Fimbriae and adhesive properties in Salmonella. J. Pathol. Bacteriol. 92:107138. Duguid, J. P., I. W. Smith, G. Dempster, and

VOL. 41, 1977 P. N. Edmunds. 1955. Nonflagellar filamentous appendages (fimbriae) and haemagglutinating activity in bacterium coli. J. Pathol. Bacteriol. 70:335-348. 36. Edebo, L. 1961. Disintegration by freeze-pressing. 1. Effects on bacteria. Acta Pathol. Microbiol. Scand. 52:300-320. 37. Edstrom, R. D., and E. G. Heath. 1965. Isolation of colitose containing oligosaccharides from the cell wall lipopolysaccharide of E. coli 111. Biochem. Biophys. Res. Commun. 21:638-643. 38. Edstrom, R. D., and E. C. Heath. 1967. The biosynthesis of cell wall lipopolysaccharide inEscherichia coli. IV. Enzymatic transfer of galactose, glucose, N-acetylglucosamine and colitose into the polymer. J. Biol. Chem. 242:3581-3588. 39. Edwards, P. R., and W. H. Ewing. 1972. Identification of Enterobacteriaceae, 3rd ed. Burgess Publishing Co., Minneapolis, Minn. 40. Erbing, C., S. Svensson, and S. Hammarstrom. 1975. Structural studies on the 0specific side-chains of the cell-wall lipopolysaccharide from E. coli 075. Carbohydr. Res. 44:259-265. 41. Evans, D. E., R. P. Silver, D. J. Evans, D. G. Chase, and S. L. Gorbach. 1975. Plasmid controlled colonization factor associated with virulence in Escherichia coli enterotoxigenic for humans. Infect. Immun. 12:656-667. 42. Ewing, W. H. 1953. Serological relationships between Shigella and coliform cultures. J. Bacteriol. 66:333-340. 43. Ewing, W. H., and B. R. Davis. 1961. The 0 antigen groups of Escherichia coli cultures from various sources. Center for Disease Control, Atlanta, Ga. 44. Ewing, W. H., and F. Kauffmann. 1950. New coli 0-antigen group. Public Health Rep. 65:1341-1343. 45. Ewing, W. H., B. R. Davis, and T. S. Montague. 1963. Studies on the occurrence of Escherichia coli serotypes associated with diarrheal disease. Center for Disease Control, Atlanta, Ga. 46. Ewing, W. H., M. C. Huchs, and M. W. Taylor. 1952. Interrelationship of certain Shigella and Escherichia cultures. J. Bacteriol. 63:319-325. 47. Ewing, W. H., K. E. Tanner, and H. W. Tatum. 1955. New serotype of Escherichia coli associated with infantile diarrhea. Public Health Rep. 70:107-114. 48. Ewing, W. H., K. E. Tanner, and H. W. Tatum. 1956. Investigation of Escherichia coli 0 group 18 serotypes isolated from cases of infantile diarrhea. Public Health Lab. 14: 106-115. 49. Ewing, W. H., H. W. Tatum, and B. R. Davis. 1958. Escherichia coli serotypes associated with edema disease of swine. Cornell Vet. 48:201-206. 50. Ewing, W. H., H. W. Tatum, B. R. Davis, and R. W. Reavis. 1956. Studies on the serology

0 AND K ANTIGENS OF E. COI

701

of the Escherichia coli group. Center for Disease Control, Atlanta, Ga. 51. Fehmel, F., U. Feige, H. Niemann, and S. Stirm. 1975. Escherichia coli capsule bacteriophages. VII. Bacteriophage 29-host capsular polysaccharide interactions. J. Virol. 16:591-601. 52. Fey, H. 1955. Serologische, biochemische und biologische Untersuchungen an Stammen aus boviner Colimastitis mit spezieller Beruicksichtigung der Coli-Sauglingsenteritis. Ergeb. Hyg. Bakteriol. Immunitaetsforsch. Exp. Ther. 29:394-474. 53. Fey, H. 1957. Bakteriologie und Serologie der Colisepsis des Kalbes. I. Serologische und biochemische Untersuchungen. Zentralbl. Veterinaermed. 4:309-318. 54. Fey, H. 1957. Bakteriologie und Serologie der Colisepsis des Kalbes. II. Die Bedeutung des Colityps 078:K80B fur die Kilber. Zentralbl. Veterinaermed. 4:447-458. 55. Fiedler, F., and L. Glaser. 1974. The synthesis of polyribitol phosphate polymerase and lipoteichoic acid carrier. J. Biol. Chem. 249:2684-2689. 56. Fiedler, F., and L. Glaser. 1974. The synthesis of polyribitol phosphate. II. On the mechanism of polyribitol phosphate polymerase. J. Biol. Chem. 249:2690-2695. 57. Frantzen, A. 1950. Kulturelle og serologiske undersogelser over nogle grupper af Enterobacteriaceae, p. 1-139. E. Munksgaard, Copenhagen. 58. Freeman, G. G. 1942. The preparation and properties of a specific polysaccharide from Bacterium typhosum Ty 2. Biochem. J. 36: 340-355. 59. Freer, J. A., and M. R. J. Salton. 1971. The anatomy and chemistry of gram-negative cell envelopes, p. 67-126. In G. Weinbaum, S. Kadis, and S. J. Ajl (ed.), Microbial toxins, vol. IV. Academic Press Inc., London. 60. Furowicz, A. J., and F. Orskov. 1972. Two new Escherichia coli 0 antigens, 0150 and 0157, and one new K antigen, K93, in strains isolated from veterinary diseases. Acta Pathol. Microbiol. Scand. Sect. B 80:441-444. 61. Galanos, C., 0. Luderitz, and 0. Westphal. 1969. A new method for the extraction of R lipopolysaccharides. Eur. J. Biochem. 9:245249. 61a. Galanos, C., and 0. Luderitz. 1975. Electrodialysis of lipopolysaccharides and their conversion to uniform salt forms. Eur. J. Biochem. 54:603-610. 62. Ganguli, L. 1970. Serological grouping of Escherichia coli in bacteriuria of pregnancy. J. Med. Microbiol. 3:201-208. 63. Garegg, P. J., B. Lindberg, T. Onn, and T. Holme. 1971. Structural studies on the Mantigen from two mucoid mutants of Salmonella typhimurium. Acta Chem. Scand. 25:1185-1194. 64. Garegg, P. J., B. Lindberg, T. Onn, and I. W. Sutherland. 1971. Comparative structural

702

ORSKOV ET AL.

studies on the M-antigen from Salmonella typhimurium, Escherichia coli and Aerobacter cloacae. Acta Chem. Scand. 25:2103-2108. 65. Gay, C. C. 1965. Escherichia coli and neonatal disease of calves. Bacteriol. Rev. 29:75-101. 66. Ghuysen, J. M., I. L. Strominger, and D. J. Tipper. 1968. Bacterial cell walls. Compr. Biochem. 26A:53-104. 67. Gillies, R. R., and J. P. Duguid. 1958. The fimbrial antigens of Shigella flexneri. J. Hyg. 56:303-318. 68. Girard, R., and A. M. Staub. 1974. Etude immunochimique de deux polysaccharides extraits de Salmonella johannesburg 5.58 souche sauvage et de Salmonella johannesburg 5.58 convertie par la phage. Carbohydr. Res. 37:127-144. 69. Glantz, P. J. 1968. Serological identification of Escherichia coli. Pennsylvania State Univ. Press, University Park. 70. Glantz, P. J., H. W. Dunne, C. E. Heist, and J. F. Hokanson. 1959. Bacteriological and serological studies of Escherichia coli serotypes associated with calf scours. Pennsylvania State University Agricultural Experiment Station Bulletin 645, p. 22. Pennsylvania State Univ. Press, University Park. 70a. Glantz, P. J., H. Rothenbacher, and J. F. Hokanson. 1968. Escherichia coli isolated from dams and their calves. Am. J. Vet. Res. 29:1561-1566. 71. Glynn, A. A., and C. J. Howard. 1970. The sensitivity to complement of strains of Escherichia coli related to their K antigens. Immunology 18:331-346. 72. Gmeiner, J. 1975. The isolation of two different lipopolysaccharide fractions from various Proteus mirabilis strains. Eur. J. Biochem. 58:621-626. 73. Goebel, W. F. 1963. Colanic acid. Proc. Natl. Acad. Sci. U.S.A. 49:464-471. 74. Golubeva, I. V., A. T. Goneim, I. N. Ulisko, and N. I. Lysenko. 1975. Serotypes of Escherichia coli isolated from patients with sepsis following illegal abortion. Zh. Mikrobiol. Epidemiol. Immunobiol. 52:143-144. 75. Grados, O., and W. H. Ewing. 1970. Antigenic relationship between Escherichia coli and Neisseria meningitidis. J. Infect. Dis. 122:100-103. 76. Greenwood, B. M., H. C. Whittle, and 0. Dominic-Rajkovic. 1971. Countercurrent immunoelectrophoresis in diagnosis of meningococcal infections. Lancet ii:519-521. 77. Gromska, W., and H. Mayer. 1976. The linkage of lysine in the O-specific chains of Proteus mirabilis 1959. Eur. J. Biochem. 62:391-399. 78. Gronroos, J. A. 1957. Investigations on Escherichia coli 0 groups 1-25 and 78 and serotypes 26:B6, 55:B5, 86:B7, 111:B4, 125:B15 and 126:B16. Ann. Med. Exp. Biol. Fenn. Suppl. 35:1-35. 79. Gronroos, J. A. 1965. Ueber das Vorkommen der Escherichia coli 1-25 in Proben aus enteralen und extraenteralen Infektionen.

BACTERIOL. REV. Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. 196:22-29. 79a.Gross, R. J., B. Rowe, A. Henderson, M. E. Byatt, and J. C. MacLaurin. 1976. A new Escherichia coli O-group, 0159, associated with outbreaks of enteritis in infants. Scand. J. Infect. Dis. 8:195-198. 80. Gruneberg, R. N., and K. A. Bettelheim. 1969. Geographical variation in serological types of urinary Escherichia coli. J. Med. Microbiol. 2:219-224. 81. Gruneberg, R. N., D. A. Leigh, and W. Brumfitt. 1968. Escherichia coli serotypes in urinary infection: studies in domiciliary, antenatal and hospital practice, p. 68-79. In F. O'Grady and W. Brumfitt (ed.), Urinary tract infection. Oxford University Press, London. 82. Guinee, P. A. M., C. M. Agterberg, and W. H. Jansen. 1972. Escherichia coli 0 antigen typing by means of a mechanized microtechnique. Appl. Microbiol. 24:127-131. 83. van Haeringen, H. 1974. Escherichia coli infections in piglets. Thesis, Drukkerij Elinkwijk, Utrecht, Netherlands, p. 1-67. 84. Hammerling, G., 0. Luderitz, 0. Westphal, and H. Makela. 1971. Structural investigations on the core oligosaccharide of Escherichia coli 0100. Eur. J. Biochem. 22:331344. 85. Hanessian, S., and T. Haskell. 1964. Structural studies on staphylococcal polysaccharide antigen. J. Biol. Chem. 239:2758-2762. 86. Harry, E. G. 1964. A study of 119 outbreaks of coli-septicaemia in broiler flocks. Vet. Rec. 76:443-449. 87. Harry, E. G., and L. A. Helmsley. 1965. The association between the presence of septicaemia strains of Escherichia coli in the respiratory and intestinal tracts of chickens and the occurrence of coli septicaemia. Vet. Rec. 77:35-40. 88. Hartley, C. L., K. Howe, A. H. Linton, K. B. Linton, and M. H. Richmond. 1975. Distribution of R plasmids among the O-antigen types of Escherichia coli isolated from human and animal sources. Antimicrob. Agents Chemother. 8:122-131. 89. Heidelberger, M. 1960. Structure and immunological specificity of polysaccharides. Fortschr. Chem. Org. Naturst. 18:503-536. 90. Heidelberger, M., and W. Nimmich. 1976. Immunochemical relationships between bacteria belonging to two separate families: Pneumococci and Klebsiella. Immunochemistry 13:67-80. 91. Heidelberger, M., K. Jann, B. Jann, F. Orskov, I. Orskov, and 0. Westphal. 1968. Relations between structures of three K polysaccharides of Escherichia coli and crossreactivity in antipneumococcal sera. J. Bacteriol. 95:2415-2417. 92. Heller, E. D., and M. Perek. 1968. Pathogenic Escherichia coli strains prevalent in poultry flocks in Israel. Br. Vet. J. 124:509-513.

VOL. 41, 1977 93. Heller, E. D., and H. Williams Smith. 1973. The incidence of antibiotic resistance and other characteristics amongst Escherichia coli strains causing fatal infection in chickens: the utilization of these characteristics to study the epidemiology of the infection. J. Hyg. 71:771-781. 94. Heller, S. J., and C. L. Villemez. 1972. Interaction of soluble glucosyl and mannosyl transferase enzyme activities in the synthesis of a glucomannan. Biochem. J. 129:645-655. 95. Henriksen, S. D. 1949. Studies in mucoidEscherichia coli. II. Specificity of the mucoid antigen (M antigen). Acta Pathol. Microbiol. Scand. 26:903-916. 96. Henriksen, S. D. 1950. Cross reacting M-antigens in E. coli and S. paratyphi B. Acta Pathol. Microbiol. Scand. 27:107-109. 97. Henriksen, S. D. 1954. Studies on the Klebsiella group (Kauffmann). IV. Cross-reactions of Klebsiella types 8, 11, 21, 26 and 35 and the M antigen of Escherichia coli. Acta Pathol. Microbiol. Scand. 34:271-275. 98. Heydanek, M. A., W. A. Struve, and F. C. Neuhaus. 1969. On the initial stage of peptidoglycan synthesis. III. Kinetics and uncoupling of phospho-N-acetylmuramyl-pentapeptide translocase (uridine 5' phosphate). Biochemistry 8:1214-1221. 99. Higashi, Y., I. L. Strominger, and C. C. Sweeley. 1967. Structure of a lipid intermediate in cell wall peptidoglycan synthesis: a derivative of a C55 isoprenoid alcohol. Proc. Natl. Acad. Sci. U.S.A. 57:1878-1884. 100. Hirata, A. A., F. C. McIntire, P. I. Terasaki, and K. K. Mittal. 1973. Cross reactions between human transplantation antigens and bacterial lipopolysaccharides. Transplantation 15:445-446. 101. Holmgren, J., and L. A. Hanson. 1969. Immunodiffusion studies of Escherichia coli. 2. Characterization of antigens used for passive haemagglutination with an 06 strain as model. Acta Pathol. Microbiol. Scand. 77:727-733. 102. Holmgren, J., G. Eggertson, L. A. Hanson, and K. Lincoln. 1969. Immunodiffusion studies on Escherichia coli. 1. Identification of 0, K and H antigens in an 06 strain. Acta Pathol. Microbiol. Scand. 76:304-318. 103. Holmgren, J., L. A. Hanson, S. E. Holm, and B. Kaijser. 1971. An antigenic relationship between kidney and certain E. coli strains. Int. Arch. Allergy Appl. Immunol. 41:463474. 104. Houwink, A. L., and W. van Iterson. 1950. Electron microscopical observations on bacterial cytology. Biochim. Biophys. Acta 5:1044. 104a. Howe, K., and A. H. Linton. 1976. The distribution of 0 antigen types of E. coli in normal calves, compared with man and their R plasmid carriage. J. Appl. Bacteriol. 40:317-330. 104b. Howe, K., A. H. Linton, and A. D. Osborne. 1976. A longitudinal study of Escherichia coli


703

in cows and calves with special reference to the distribution of 0 antigen types and antibiotic resistance. J. Appl. Bacteriol. 40:331340. 105. Hungerer, D., K. Jann, B. Jann, F. Orskov, and I. Orskov. 1967. Immunochemistry of K antigens of Escherichia coli. 4. The K antigen of E. coli 09:K30:H12. Eur. J. Biochem. 2:115-126. 105a.Isaacson, R. E. 1977. K99 surface antigen of Escherichia coli: purification and partial characterization. Infect. Immun. 15:272-279. 105b. Isaacson, R. E., B. Nagy, and H. W. Moon. Colonization of porcine small intestine by Escherichia coli: colonization and adhesion factors of pigenteropathogens which lack K88. J. Infect. Dis. 135:531-539. 106. Jann, B., K. Reske, and K. Jann. 1975. Heterogeneity of lipopolysaccharide. Analysis of polysaccharide chain lengths by sodium do-

decyl-sulfate-polyacrylamide gel electrophoresis. Eur. J. Biochem. 60:239-246. 107. Jann, B., K. Jann, G. Schmidt, I. Orskov, and F. Orskov. 1970. Immunochemical studies of polysaccharide surface antigens of E. coli O100:K?(B):H2. Eur. J. Biochem. 15:29-39. 108. Jann, B., K. Jann, G. Schmidt, F. 0rskov, and I. 0rskov. 1971. Comparative immunochemical studies of the surface antigens of Escherichia coli strains 08:K87(B?):H19 and (032):K87(B?):H45. Eur. J. Biochem. 23:515522. 109. Jann, K. 1969. Neuraminsiurehaltige Polysaccharide in Escherichia coli. Z. Physiol. Chem. 350:666. 110. Jann, K. 1975. Immunchemie der Oberflachenantigene von E. coli. Arzneim. Forsch. 25:989. 111. Jann, K., and B. Jann. 1976. Bacterial polysaccharide antigens, p. 247-285. In I. W. Sutherland (ed.), Surface carbohydrates of the procaryotic cell. Academic Press Inc., London. 112. Jann, K., and 0. Westphal. 1975. Microbial polysaccharides, p. 1-125. In M. Sela (ed.), The antigens, vol. III. Academic Press Inc., London. 113. Jann, K., B. Jann, F. Orskov, and I. Orskov. 1966. Immunchemische Untersuchungen an K-Antigenen von Escherichia coli. III. Isolierung und Untersuchung der chemischen Struktur des sauren Polysaccharids aus E. coli 0141:K85(B):H4 (K85-Antigen). Biochem. Z. 346:368-385. 114. Jann, K., G. Schmidt, B. Wallenfels, and E. Freund-Molbert. 1971. Isolation and characterization of Escherichia coli bacteriophage fQ specific for E. coli strains belonging to sero-group 08. J. Gen. Microbiol. 67:289-297. 115. Jann, K., B. Jann, F. 0rskov, I. Orskov, and 0. Westphal. 1965. Immunchemische Untersuchungen an K Antigenen von Escherichia coli. II. Das K Antigen von E. coli 08:K42(A):H-. Biochem. Z. 342:1-22. 116. Jann, K., B. Jann, K. F. Schneider, F. 0rskov, and I. Orskov. 1968. Immunochemistry of K

704

BACTERIOL. PRBV.

ORSKOV ET AL.

antigen of Escherichia coli. 5. The K antigen of E. coli 08:K27(A):H-. Eur. J. Biochem. 5:456-465. 117. Jensen, C. 0. 1892-93. Om den infektiose Kalvediarrhoe og dens Aarsag. Maanedsskrift Dyrlaeger 4:140-145. 118. Jones, G. W., and J. M. Rutter. 1974. The association of K88 antigen with haemagglutinating activity in porcine strains ofE. coli. J. Gen. Microbiol. 84:135-144. 119. Jones, R. T., D. E. Koeltzow, and B. A. D. Stocker. 1975. Genetic transfer of Salmonella typhimurium and Escherichia coli lipopolysaccharide antigens to Escherichia coli K-12. J. Bacteriol. 111:758-770. 120. Kaeckenbeeck, A., and J. Thomas. 1960. A propos des serotypes colibacillaires dans la diarrh6e des veaux. Determination des antigenes somatiques (0). Ann. Med. Vet. 5:232-239. 121. Kaoser, B. 1972. E. coli 0 and K antigens and protective antibodies in relation to urinary tract infection. Thesis, Goteborg University, Gothenburg, Sweden. 121a. Kaiser, B. 1977. A simple method for typing of acidic polysaccharide K antigens ofE. coli. FEMS, Microbiol. Lett. 1:285-288. 121b.Kajser, B., L. A. Hanson, U. Jodel, G. LidinJanson and J. B. Robbins. 1977. Frequency of E. coli K antigens in urinary-tract infections in children. Lancet i:663-664. 122. Kampelmacher, E. H. 1959. On antigenic 0 relationships between the groups Salmonella, Arizona, Escherichia and Shigella. Antonie van Leeuwenhoek J. Microbiol. Serol. 25:289-324. 123. Kang, S., and A. Markovitz. 1967. Induction of capsular polysaccharide synthesis by p-fluorophenylalanine in Escherichia coli wild type and strains with altered phenylalanyl soluble ribonucleic acid synthetase. J. Bacteriol. 93:584-591. 124. Kasper, D. L., J. L. Winkelhage, W. D. Zollinger, B. Brandt, and M. S. Artenstein. 1973. Immunological similarity between polysaccharide antigens of Escherichia coli 07:K1(L):NM and group B Nei88eria meningitidis. J. Immunol. 110:262-268. 125. Kauffmann, F. 1935. Ueber einen neuen serologischen Formenwechsel der Typhusbacillen. Z. Hyg. Infektionskr. 116:617-651. 126. Kauffmann, F. 1936. Untersuchungen uber die Korperantigene in der Salmonella-Gruppe. Z. Hyg. Infektionskr. 117:778-791. 127. Kauffmann, F. 1943. Ueber neue thermolabile K6rperantigene der Colibakterien. Acta Pathol. Microbiol. Scand. 20:21-44. 128. Kaufmann, F. 1944. Zur Serologie der ColiGruppe. Acta Pathol. Microbiol. Scand. 21:20-45. 129. Kauffmann, F. 1949. On the serology of the Klebsiella group. Acta Pathol. Microbiol. Scand. 26:381-406. 130. Kauffmann, F. 1954. Enterobacteriaceae, 2nd ed. E. Munksgaard, Copenhagen. 131. Kauffmann, F. 1961. Die Bakteriologie der

132. 133.

134.

135.

136.

Salmonella species. E. Munksgaard, Copenhagen. Kauffmann, F. 1966. The bacteriology ofEnterobacteriaceae. E. Munksgaard, Copenhagen. Kauffmann, F., and A. Dupont. 1950. Escherichia strains from infantile epidemic gastroenteritis. Acta Pathol. Microbiol. Scand. 27:552-564. Kauffmann, F., and F. Orskov. 1956. Die Bakteriologie derEscherichia coli-enteritis, p. 141. In A. Adam (ed.), Sauglings-Enteritis. Georg Thieme Verlag, Stuttgart. Kauffmann, F., and G. Vahlne. 1945. Ueber die Bedeutung des serologischen Formenwechsels fMr die Bakteriophagenwirkung in der Coli-Gruppe. Acta Pathol. Microbiol. Scand. 22:119-137. Kelen, A. E., S. G. Campbell, and D. A. Barnum. 1959. Studies on haemolytic Escherichia coli strains associated with oedema disease of swine. Can. J. Comp. Med. 23:216-

221. 137. Kiefer, W., G. S. Schmidt, B. Jann, and K. Jann. 1976. Genetic transfer of Salmonella 0 antigens to Escherichia coli 08. J. Gen. Microbiol. 92:311-324. 138. Knipschildt, H. E. 1945. Undersogelser over Coligruppens Serologi. Nyt Nordisk Forlag, Arnold Busck, Kobenhavn. 139. Knolle, P., and I. Orskov. 1967. The identity of the f+ antigen and the cellular receptor for RNA phage fr. Mol. Gen. Genet. 99:109-114. 140. Knox, K. W., and A. J. Wicken. 1973. Immunological properties of teichoic acids. Bacteriol. Rev. 37:215-257. 141. Kochetkov, N. K., B. A. Dmitriev, L. V. Backinowski, and V. L. Lvov. 1975. New sugars isolated from Shigella antigen lipopolysaccharides (in Russian). Bioorg. Chem. 1:12381240. 142. Kopmann, H. J., and K. Jann. 1975. Biosynthesis of the 09 antigen of Escherichia coli. The polysaccharide component of E. coli 09:K29-. Eur. J. Biochem. 60:587-601. 143. Kramer, T. 1960. Study of hemolytic Escherichia coli strains isolated from cases of edema disease in swine. Can. J. Comp. Med. 24:289-294. 144. Kunin, C. M., and M. V. Beard. 1963. Serological studies of 0 antigens of Escherichia coli by means of the hemagglutination test. J. Bacteriol. 85:541-548. 145. Kunin, C. M., M. V. Beard, and N. Halmagyi. 1962. Evidence for a common hapten associated with endotoxin fractions of E. coli and other Enterobacteriaceae. Proc. Soc. Exp. Biol. Med. 111:160-166. 146. Lautrop, H., I. Orskov, and K. Gaarslev. 1971. Hydrogen sulphide producing variants of Escherichia coli. Acta Pathol. Microbiol. Scand. Sect. B 79:641-650. 147. Lawn, A. M., E. Meynell, G. G. Meynell, and N. Datta. 1967. Sex pili and the classification of sex factors in the Enterobacteriaceae. Nature (London) 216:343-346.

.VOL. 41, 1977 148. Lawson, C. J., C. W. McCleary, H. J. Nakada, D. A. Rees, I. W. Sutherland, and J. F. Wilkinson. 1969. Structural analysis of colanic acid from Escherichia coli by using methylation and base-catalyzed fragmentation. Comparison with polysaccharides from other bacterial sources. Biochem. J. 115:947-958. 149. Le-Ba Nhan, B. Jann, and K. Jann. 1971. Immunochemistry of K antigens of Escherichia coli. The K29 antigen of E. coli 09:K29(A):H-. Eur. J. Biochem. 21:226-234. 150. Leive, L. 1965. Release of lipopolysaccharide by EDTA of E. coli. Biochem. Biophys. Res. Commun. 21:290-296. 151. Lennarz, W. J., and M. G. Scher. 1972. Metabolism and function of polyisophenol sugar intermediates in membrane associated media. Biochim. Biophys. Acta 265:417-441. 152. Lieberman, M. M., and A. Markovitz. 1970. Depression of guanosine diphosphate-mannose pyrophosphorylase by mutations in two different regulator genes involved in capsular polysaccharide synthesis in Escherichia coli K-12. J. Bacteriol. 101:965-972. 153. Lindberg, B., J. L6nngren, and W. Nimmich. 1972. Structural studies onKlebsiella 0 group 5 lipopolysaccharides. Acta Chem. Scand. 26: 2231-2236. 154. Lovell, R. 1957. Some aetiological factors in diseases of the young. Vet. Res. 69:13141317. 155. Luderitz, O., K. Jann, and R. Wheat. 1968. Somatic and capsular antigens of gram-negative bacteria. Compr. Biochem. 26A:105228. 156. Luderitz, D., A. M. Staub, and 0. Westphal. 1966. Immunochemistry of 0 and R antigens of Salmonella and related Enterobacteriaceae. Bacteriol. Rev. 30:193-255. 157. Luderitz, O., 0. Westphal, A. M. Staub, and H. Nikaido. 1971. Isolation and chemical and immunological characterization of bacterial lipopolysaccharides, p. 145-233. In G. Weinbaum, S. Kadis, and S. J. Ajl (ed.), Microbial toxins, vol. IV. Academic Press Inc., London. 158. Luderitz, O., C. Galanos, V. Lehmann, M. Nurminen, E. Rietschel, G. Rosenfelder, M. Simon, and 0. Westphal. 1973. Lipid A, chemical structure and biological activity. J. Infect. Dis. 128(Suppl.):9-21. 159. Mabeck, C. E., F. Orskov, and I. Orskov. 1971. Studies in urinary infections. VIII. Escherichia coli O:H serotypes in recurrent infections. Acta Med. Scand. 190:279-282. 160. Mikeli, P. H., and H. Mayer. 1976. Enterobacterial common antigen. Bacteriol. Rev. 40:591-632. 161. MikelJ, P. H., M. Jahkola, and 0. Luderitz. 1970. A new gene cluster rfe concerned with biosynthesis of Salmonella lipopolysaccharide. J. Gen. Microbiol. 60:91-106. 162. Manz, J. 1971. Ueber die serologische Differenzierung von Escherichia coli Stimme vom Kalb. Thesis, Universitit Munchen, Munich.


705

162a.Marier, R., J. G. Wells, R. C. Swanson, W. Callahan, and I. J. Mehlman. 1973. An outbreak of enteropathogenic Escherichia coli foodborne disease traced to imported cheese. Lancet ii:1376-1378. 163. Markovitz, A. 1964. Regulatory mechanisms for synthesis of capsular polysaccharide in mucoid mutants of Escherichia coli K12. Proc. Natl. Acad. Sci. U.S.A. 51:239-246. 164. Markovitz, A., M. M. Lieberman, and N. Roesenbaum. 1967. Depression of phosphomannose isomerase by regulator gene mutations involved in capsular polysaccharide synthesis in Escherichia coli K-12. J. Bacteriol. 94:1497-1501. 165. Markovitz, A., R. J. Sydiskis, and M. M. Lieberman. 1967. Genetic and biochemical studies on mannose-negative mutants that are deficient in phosphomannose in Escherichia coli K-12. J. Bacteriol. 94:1492-1496. 166. Mayer, H. 1969. D-MannosaminuronsiureBaustein des K7-Antigens von Escherichia coli. Eur. J. Biochem. 8:139-145. 167. Mayer, H., M. C. Rapin, G. Schmidt, and H. G. Boman. 1976. Immunochemical studies on lipopolysaccharides from wild-type and mutants of Escherichia coli K-12. Eur. J. Biochem. 66:357-358. 168. McGuire, E. J., and S. B. Binkley. 1964. The structure and chemistry of colominic acid. Biochemistry 3:247-251. 169. Moller, 0. 1948. Bacterial variation in Escherichia coli. Gleerupske Univ. Bokhandeln, Lund, Sweden. 170. Morch, E., and H. E. Knipschildt. 1944. Serological relationship between a strain of bac. coli and some pneumococcus types. Acta Pathol. Microbiol. Scand. 21:102-109. 171. Morrison, D. C., and L. Leive. 1975. Fractions of lipopolysaccharide from Escherichia coli O111:B4 prepared by two extraction procedures. J. Biol. Chem. 250:2911-2919. 172. Mfihlradt, P. 1975. Die Zelloberfliche gramnegativer Bakterien. Biol. Unserer Zeit 5:33-39. 173. Mfiller-Seitz, E., B. Jann, and K. Jann. 1968. Degradation studies on the lipopolysaccharide from E. coli 071:K?:H12. Separation and investigation of O-specific and core polysaccharides. FEBS Lett. 1:311-314. 174. Mushin, R. 1949. A new antigenic relationship among faecal bacilli due to a common /3 antigen. J. Hyg. 47:227-235. 175. Nikaido, H. 1973. Biosynthesis and assembly of lipopolysaccharide and the outer membrane layer of gram-negative cell wall, p. 131-208. In L. Leive (ed.), Microbiology series, vol. I: bacterial membranes and walls. Marcel Dekker Inc., New York. 176. Nimmich, W., and G. Korten. 1970. Chemical composition of Klebsiella lipopolysaccharides (0-antigens). Pathol. Microbiol. 36:179-184. 177. Nimmich, W., L. Girtner, E. Budde, and G. Naumann. 1975. Serotypiesierung von Escherichia coli bei Harnweginfektionen. Zen-

706

178.

179.

180.

181.

182.

183.

184. 185.

186.

187.

188. 189.

190.

191.

192.

ORSKOV ET AL. tralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. Reihe A 230:28-37. Norval, M., and I. V. Sutherland. 1969. A group of Klebsiella mutants showing temperature dependent polysaccharide synthesis. J. Gen. Microbiol. 57:369-377. Ogawa, H., A. Nakamura, and R. Sakazaki. 1968. Pathogenic properties of "enteropathogenic" Escherichia coli from diarrheal children and adults. Jpn. J. Med. Sci. Biol. 21:333-349. Old, D. C., and S. B. Payne. 1971. Antigens of type-2 fimbriae of Salmonellae: "cross reacting material" (CRM) of type-1 fimbriae. J. Med. Microbiol. 4:215-225. Orskov, F. 1951. On the occurrence of E. coli belonging to 0 group 26 in cases of infantile diarrhoea and white scours. Acta Pathol. Microbiol. Scand. 29:373-378. 0rskov, F. 1952. Antigenic relationships between 0 groups 1-112 of E. coli and Wramby's 0 groups 26w-43w. Acta Pathol. Microbiol. Scand. 31:51-56. 0rskov, F. 1952. Antigenic relationships between H antigens 1-22 of E. coli and Wramby's H antigens 23w-36w. Acta Pathol. Microbiol. Scand. 32:241-244. Orskov, F. 1954. Studies on a number of Escherichia coli strains belonging to 0 group 86. Acta Pathol. Microbiol. Scand. 35:179-186. 0rskov, F. 1956. Escherichia coli. Om typebestemmelse, specielt med henblik pA stammer fra infantil diarrhoe. Thesis, Nyt Nordisk Forlag, Arnold Busck, Copenhagen. Orskov, F. 1956. Escherichia coli strains isolated from cases of infantile diarrhoea and from healthy infants. Acta Pathol. Microbiol. Scand. 39:137-146. 0rskov, F. 1976. Agar electrophoresis combined with second dimensional Cetavlon precipitation. A new method for demonstration of acidic polysaccharide K antigens. Acta Pathol. Microbiol. Scand. Sect. B 84:319-320. Orskov, F., and I. Orskov. 1962. Behaviour of E. coli antigens in sexual recombination. Acta Pathol. Microbiol. Scand. 55:99-109. Orskov, F., and I. Orskov. 1969. Staining of Escherichia coli K88 antigen and other fimbrial antigens for light microscopy. J. Gen. Microbiol. 55:21. 0rskov, F., and I. Orskov. 1972. Immunoelectrophoretic patterns of extracts from Escherichia coli 0 antigen test strains 01 to 0157. Examinations in homologous 0 and OK sera. Acta Pathol. Microbiol. Scand. Sect. B 80:905-910. 0rskov, F., and I. Orskov. 1975. Escherichia coli O:H serotypes isolated from human blood. Acta Pathol. Microbiol. Scand. Sect. B 83:595-600. Orskov, F., and I. Orskov. 1977. Serotyping of Enterobacteriaceae, with special emphasis on K antigen determination. In T. Bergan and J. R. Norris (ed.), Methods in microbiology, vol. 10. Academic Press Inc., London, in

BACTEMIOL. Rzv. press. 193. 0rskov, F., I. Orskov, and A. J. Furowicz. 1972. Four new Escherichia coli 0 antigens, 0148, 0151, 0152, 0153, and one new H antigen, H50, found in strains isolated from enteric diseases in man. Acta Pathol. Microbiol. Scand. Sect. B 80:435-440. 194. Orskov, F., I. 0rskov, B. Jann, and K. Jann. 1971. Immunoelectrophoretic patterns of extracts from all Escherichia coli 0 and K antigen test strains. Correlation with pathogenicity. Acta Pathol. Microbiol. Scand. Sect. B 79:142-152. 195. Orskov, F., I. Orskov, T. A. Rees, and K. Sahab. 1960. Two new E. coli 0 antigens: 0141 and 0142, and two new E. coli K antigens: K85 and K86. Acta Pathol. Microbiol. Scand. 48:48-50. 196. Orskov, F., I. Orskov, D. J. Evans, Jr., R. B. Sack, D. A. Sack, and T. Wadstr6m. 1976. Special Escherichia coli serotypes among enterotoxigenic strains from diarrhoea in adults and children. Med. Microbiol. Immunol. 162:73-80. 197. Orskov, F., I. 0rskov, B. Jann, K. Jann, E. Muller-Seitz, and 0. Westphal. 1967. Immunochemistry of Escherichia coli 0 antigens. Acta Pathol. Microbiol. Scand. 71:339-358. 198. Orskov, I. 1954. 0 antigens in the Klebsiella group. Acta Pathol. Microbiol. Scand. 34:145-156. 199. 0rskov, I., and K. Nyman. 1974. Genetic mapping of the antigenic determinants of two polysaccharide K antigens, K1O and K54, in Escherichia coli. J. Bacteriol. 120:43-51. 200. 0rskov, I., and F. 0rskov. 1960. An antigen termed f+ occurring in F+ E. coli strains. Acta Pathol. Microbiol. Scand. 48:37-46. 201. Orskov, I., and F. 0rskov. 1960. The H antigen of the "K12" strain. A new E. coli H antigen: H48. Acta Pathol. Microbiol. Scand. 48:47. 202. Orskov, I., and F. Orskov. 1966. Episome-carried surface antigen K88 of Escherichia coli. I. Transmission of the determinant of the K88 antigen and influence on the transfer of chromosomal markers. J. Bacteriol. 91:6975. 203. 0rskov, I., and F. 0rskov. 1968. What is the nature of E. coli K antigens? An attempt to analyse the K antigens by various serological methods. Arch. Immunol. Ther. Exp. (Engl. Transl.) 16:387-391. 204. Orskov, I., and F. Orskov. 1970. The K antigens ofEscherichia coli. Re-examination and re-evaluation of the nature of L antigens. Acta Pathol. Microbiol. Scand. Sect. B 78:593-604. 205. Orskov, I., and F. Orskov. 1976. Five new Escherichia coli K antigens, K95, K96, K97, K98 and K100. Acta Pathol. Microbiol. Scand. Sect. B 84:321-325. 205a.Orskov, I., and F. Orskov. 1977. Special O:K:H serotypes among enterotoxigenic E. coli strains from diarrhoea in adults and children. Occurrence of the CF (colonization fac-


VOL. 41, 1977 tor) antigen and of hemagglutinating abilitieh. Med. Microbiol. Immunol. 163:99-110. 206. Orskov, I., F. Orskov, and B. Rowe. 1973. Three new Escherichia coli 0 antigens, 0154, 0155, 0156 and one new K antigen, K94. Acta Pathol. Microbiol. Scand. Sect. B 81:59-64. 207. Orskov, I., F. Orskov, and B. Rowe. 1975. Four new Eacherichia coli 0 antigens, 0158, 0159, 0160 and 0161, and two new H antigens, H53 and H54. Acta Pathol. Microbiol. Scand. Sect. B 83:116-120. 208. Orskov, I., V. Sharma, and F. Orskov. 1976. Genetic mapping of the K1 and K4 antigens (L) of Escherichia coli. Non-allelism of K(L) antigens with K antigens of 08:K27(A), 08:K8(L) and 09:K57(B). Acta Pathol. Microbiol. Scand. Sect. B 84:125-131. 209. Orskov, I., F. Orskov, K. A. Bettelheim, and M. E. Chandler. 1975. Two new Escherichia coli 0 antigens, 0162 and 0163, and one new H antigen, H56. Withdrawal of H antigen H50. Acta Pathol. Microbiol. Scand. Sect. B 83:121-124. 210. 0rskov, I., F. Orskov, B. Jann, and K. Jann. 1963. Acidic polysaccharide antigens of a new type from E. coli capsules. Nature (London) 200:144-146. 211. 0rskov, I., F. (rskov, W. J. Sojka, and J. M. Leach. 1961. Simultaneous occurrence of E. coli B and L antigens in strains from diseased swine. Acta Pathol. Microbiol. Scand. 53:404-422. 212. Orskov, I., F. Orskov, W. J. Sojka, and H. W. Smith. 1975. The establishment of K99, a thermolabile transmissible Escherichia coli K antigen, previously called "Kco," possessed by calf and lamb enteropathogenic strains. Acta Pathol. Microbiol. Scand. Sect. B 83:31-36. 213. 0rskov, I., F. Orskov, W. J. Sojka, and W. Wittig. 1964. K antigens K88ab(L) and K88ac(L) in E. coli. Acta Pathol. Microbiol. Scand. 62:439-447. 214. Orskov, I., F. 0rskov, W. Wittig, and E. J. Sweeney. 1969. A new E. coli serotype 0149:K91(B), K88ac(L):HlO isolated from diseased swine. Acta Pathol. Microbiol. Scand. 75:491-498. 215. Osborn, M. J., and L. I. Rothfield. 1971. Biosynthesis of the core region of lipopolysaccharide, p. 331-350. In G. Weinbaum, S. Kadis, and S. J. Ajl (ed.), Microbial toxins, vol. IV. Academic Press Inc., London. 216. Osborn, M. J., I. E. Gander, E. Parisi, and J. Carson. 1972. Mechanism of assembly of the outer membrane of Salmonella typhimurium. Isolation and characterization of cytoplasmic and outer membrane. J. Biol. Chem. 247:3962-3972. 217. Osborn, M. J., M. A. Cynkin, I. M. Gilberg, L. Muller, and M. Singh. 1972. Synthesis of bacterial 0-antigens. Methods Enzymol. 28:583-601. 218. Ouchterlony, 0. 1949. In vitro method for test-

707

ing the toxin-producing capacity of diphtheria bacteria. Acta Pathol. Microbiol. Scand. 26:516-524. 219. Ouchterlony, 0. 1958. Diffusion-in-gel methods for immunological analysis. Prog. Allergy 5:1-78. 220. Parodi, A. J., N. H. Behrens, L. F. Leloir, and H. Carminiatti. 1972. The role of polyprenolbound saccharides as intermediates in glycoprotein synthesis in liver. Proc. Natl. Acad. Sci. U.S.A. 69:3268-3272. 221. Pless, D. D., and W. J. Lennarz. 1973. A lipidlinked oligosaccharide intermediate in glycoprotein synthesis. Characterization of [man-

222.

223.

224.

225.

14C]glycoproteins labelled from [man-14C]oligosacharide lipid and [GDP-14C]man. J. Biol. Chem. 250:7014-7019. Prehm, P., B. Jann, and K. Jann. 1976. The 09 antigen of Escherichia coli structure of the polysaccharide chain. Eur. J. Biochem. 67:53-56. Prehm, P., S. Stirm, B. Jann, and K. Jann. 1975. Cell wall lipopolysaccharide from Escherichia coli B. Eur. J. Biochem. 56:4155. Prehm, P., B. Jann, K. Jann, G. Schmidt, and S. Stirm. 1976. On a bacteriophage T3 and T4 receptor region within the cell wall lipopolysaccharide of Escherichia coli. J. Mol. Biol. 101:277-281. Prehm, P., S. Stirm, B. Jann, K. Jann, and H. G. Boman. 1976. Cell wall lipopolysaccharides of ampicillin resistant mutants of Escherichia coli K-12. Eur. J. Biochem. 66:369-

377. 225a. Prehm, P., G. Schmidt, B. Jann, and K. Jann. 1976. The cell wall lipopolysaccharide of Escherichia coli K-12. Structure and acceptor site for O-antigen and other substituents. Eur. J. Biochem. 70:171-177. 226. Radke, K. L., and E. C. Siegel. 1971. Mutation preventing capsular polysaccharide synthesis in Escherichia coli K-12 and its effect on bacteriophage resistance. J. Bacteriol. 106:432-437. 227. Rantz, L. A. 1962. Serological grouping of Each. coli: study in urinsry tract infection. Arch. Intern. Med. 109:37-42. 228. Rees, T. A. 1958. Studies on E. coli of animal origin. I. E. coli from natural outbreaks of colibacillosis of calves. J. Comp. Pathol. 68:388-398. 229. Refai, M., and R. Rohde. 1975. Neue serologische Untersuchungen fiber O-Antigenbeziehungen zwischen Escherichia coli, Salmonella, Arizona und Citobacer. Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt 1 Orig. Reihe A 233:171-179. 230. Reske, K., and K. Jann. 1972. The 08 antigen of Escherichia coli. Structure of the polysaccharide chain. Eur. J. Biochem. 31:320-328. 231. Robbins, J. B., G. H. McCracken, E. C. Gotschlich, F. 0rskov, I. Orskov, and L. A. Hanson. 1974. Escherichia coli K1 cap sular polysaccharide associated with neo-

708

0RSKOV ET AL.

natal meningitis. N. Engl. J. Med. 290:12161220. 232. Robbins, J. B., R. L. Myerowitz, J. K. Whisnant, M. Argaman, R. Schneerson, Z. T. Handzel, and E. C. Gotschlich. 1972. Enteric bacteria cross-reactive with Neisseria meningitidis groups A and C and Diplococcus pneumoniae types I and III. Infect. Immun. 6:651656. 233. Robbins, J. B., R. Schneerson, I. C. Parke, T.Y. Liu, Z. T. Handzel, I. Orskov, and F. 0rskov. 1976. Escherichia coli 075:Kfl47:H5 cross-reactive with the capsular polysaccharide ofHaemophilus influenzae type B: studies of the immunochemical relationship and prospects for heteroimmunization in humans to prevent septicemic disease caused by H. influenzae type B, p. 103-120. In R. F. Beers, Jr., and E. G. Bassett (ed.), The role of immunological factors in infectious, allergic, and autoimmune processes. Raven Press, New York. 234. Robbins, J. B., R. Schneerson, T. Liu, M. S. Schiffer, G. Schiffmann, R. L. Myerowitz, G. H. McCracken, Jr., I. Orskov, and F. 0rskov. 1975. Cross-reacting bacterial antigens and immunity to disease caused by encapsulated bacteria, p. 218-241. In E. Neter and F. Milgrom (ed.), The immune system and infectious diseases. 4th International Convocation on Immunology, Buffalo, N.Y. Karger, New York. 235. Robbins, P. W., and A. Wright. 1971. Biosynthesis of 0-antigens, p. 351-368. In G. Weinbaum, S. Kadis, and S. J. Ajl (ed.), Microbial toxins, vol. IV. Academic Press Inc., London. 236. Rothfield, L., and R. W. Horne. 1967. Reassociation of purified lipopolysaccharide and phospholipid of the bacterial cell envelope: electron microscopic and monolayer studies. J. Bacteriol. 93:1705-1715. 237. Rowe, B., R. J. Gross, and M. Guiney. 1976. Antigenic relationships between Escherichia coli 0 antigens 0149 to 0163 and Shigella 0 antigens. Int. J. Syst. Bacteriol. 26:76-78. 237a.Rowe, B., R. J. Gross, R. Lindop, and R. B. Baird. 1974. A new E. coli 0 group, 0158, associated with an outbreak of infantile enteritis. J. Clin. Pathol. 27:832-835. 237b. Rowe, B., R. J. Gross, and D. P. Woodroof. 1977. Proposal to recognize serovar 145/46 as a new Escherichia coli 0 group, 0164. Int. J. Syst. Bacteriol. 27:15-18. 238. Sack, R. B. 1975. Human diarrheal disease caused by enterotoxigenic Escherichia coli. Annu. Rev. Microbiol. 29:333-353. 239. Sack, R. B., D. A. Sack, I. J. MehIman, I. Orskov, and F. Orskov. 1977. Enterotoxigenic Escherichia coli from food sources. J. Infect. Dis. 135:313-317. 240. Sakazaki, R., K. Tamura, A. Nakamura, T. Kurata, A. Gohda, and S. Takeuchi. 1974. Enteropathogenicity and enterotoxigenicity of human enteropathogenic Escherichia coli. Jpn. J. Med. Sci. Biol. 27:19-33.

BACTECRIOL. REV. 241. Sarff, L., G. H. McCracken, Jr., M. S. Schiffer, M. P. Glode, J. B. Robbins, I. 0rskov, and F. 0rskov. 1975. Epidemiology of Escherichia coli K1 in healthy and diseased newborns. Lancet i:1099-1104. 241a.Scheidegger, J. J. 1955. Une micromethode de l'immuno-6lectrophorese. Int. Arch. Allergy 7:103-110. 242. Scher, M., and W. J. Lennarz. 1972. Synthesis of mannan in Micrococcus lysodeiticus; lipid bound intermediates. Methods Enzymol. 28:563-571. 244. Schmidt, G. 1972. Basalstrukturen von verschiedenen Lipopolysacchariden verschiedener Enterobakteriaceen. Genetische und serologische Untersuchungen an R-Mutanten. Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. Reihe A 220:472476. 245. Schmidt, G. 1973. Genetical studies on the lipopolysaccharide structure ofEscherichia coli K12. J. Gen. Microbiol. 77:151-160. 246. Schmidt, G., I. Fromme, and H. Mayer. 1970. Immunochemical studies on core lipopolysaccharides of Enterobacteriaceae of different genera. Eur. J. Biochem. 14:357-366. 247. Schmidt, G., B. Jann, and K. Jann. 1969. Immunochemistry of R lipopolysaccharides of Escherichia coli. Different core structures in the lipopolysaccharides of 0 group 8. Eur. J. Biochem. 10:501-510. 248. Schmidt, G., B. Jann, and K. Jann. 1970. Immunochemistry of R lipopolysaccharides of Escherichia coli. Studies on R mutants with an incomplete core, derived from E. coli 08:K27. Eur. J. Biochem. 16:382-392. 249. Schmidt, G., B. Jann, and K. Jann. 1974. Genetic and immunochemical studies on Escherichia coli 014:K7:H-. Eur. J. Biochem. 42:303-309. 250. Schmidt, G., D. Mannel, H. Mayer, H. Y. Whang, and E. Neter. The role of a lipopolysaccharide gene for immunogenicity of the enterobacterial common antigen. J. Bacteriol. 126:579-586. 251. Schmidt, G., H. Mayer, and P. H. Makela 1976. The presence of rfe genes in Escherichia coli: their participation in biosynthesis of 0 antigen and enterobacterial common antigen. J. Bacteriol. 127:755-762. 251a. Schmidt, G., B. Jann, K. Jann, I. 0rskov, and F. Orskov. 1977. Genetic determinants of the synthesis of the polysaccharide capsular antigen K(A)27 ofEscherichia coli. J. Gen. Microbiol. 100:355-361. 252. Schneerson, R., M. Bradshaw, J. K. Whisnant, R. L. Myerowitz, J. C. Parke, Jr., and J. B. Robbins. 1972. AnEscherichia coli antigen cross-reactive with capsular polysaccharide of Haemophilus influenzae type b. Occurrence among known serotypes and immunochemical and biologic properties of E. coli antisera toward H. influenzae type b. J. Immunol. 108:1551-1562. 253. Schoenaers, F., and A. Kaeckenbeeck. 1974.

VOL. 41, 1977

Septicemie colibacillaire du veau nouveaune. Antigbnes somatiques des colibacilles isoles en Belgique de 1960 a 1966. Leur r6le 6tiologique. Ann. Med. Vet. 118:1-7. 254. Scott, J. E. 1960. Aliphatic ammonium salts in the assay of acidic polysaccharides from tissues. Methods Biochem. Anal. 8:145-197. 255. Sears, H. J., I. Brownley, and F. E. Uchiyame. 1950. Persistence of individual strains of Escherichia coli in the intestinal tract of man. J. Bacteriol. 59:293-301. 256. Sedlak, J., and H. Rische. 1961. Enterobacteriaceae Infektionen, p. 279-310. G. Thieme, Leipzig. 257. Seeliger, H. 1955. Antigenverwandschaft zwischen Shigella flexneri Typ 5 und einer neuen Escherichia coli-O-Gruppe 129. Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. Reihe A 163:7-12. 258. Sela, M., T. Schechter, B. Schechter, and F. Bozek. 1961. Antibodies to sequential and conformational determinants. Cold Spring Harbor Symp. Quant. Biol. 32:537-542. 259. Seltmann, G. 1971. Untersuchungen zur Antigenstruktur von Shigellen. 4. Mitteilung: Isolierung und Reinigung eines thermolabilen Antigens von Sh. sonnei. Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. Reihe A 219:324-335. 260. Senijen, G., I. 0rakov, and F. 0rskov. 1977. K antigen determination of Escherichia coli by counter-current immunoelectrophoresis (CIE). Acta Pathol. Microbiol. Scand. Sect. B 85:103-107. 261. Sereny, B. 1967. Experimental kerato conjunctivitis shigellosa. Acta Microbiol. Acad. Sci. Hung. 4:367-376. 262. Shands, J. W. 1971. The physical structure of bacterial lipopolysaccharides, p. 127-144. In G. Weinbaum, S. Kadis, and S. J. Ajl (ed.), Microbial toxins, vol IV. Academic Press Inc., London. 263. Smith, H. W., and M. A. Linggood. 1971. Further observations on Escherichia coli enterotoxins with particular regard to those produced by atypical piglet strains and by calf and lamb strains: the transmissible nature of these enterotoxins and of a K antigen possessed by calf and lamb strains. J. Med. Microbiol. 5:243-250. 264. Soderlind, 0. 1965. Serological investigation of O antigen in Escherichia coli strains isolated from calves with coli septicaemia. Zentralbl. Veterinaermed. 12:13-17. 265. Soderlind, 0. 1971. Studies on Escherichia coli in pigs. I. Serological investigations. Zentralbl. Veterinaermed. Reihe B 18:569-590. 266. Soika, W. J. 1965. Escherichia coli in domestic animals and poultry, p. 1-131. Commonwealth Agricultural Bureaux, Farnham Royal, Bucks, England. 267. Sojka, W. J., and R. B. A. Carnaghan. 1961. Escherichia coli infection in poultry. Res. Vet. Sci. 2:340-352. 268. Sojka, W. J., M. K. Lloyd, and E. J. Sweeney.


709

1960. Escherichia coli serotypes associated with certain pig diseases. Res. Vet. Sci. 1:1727. 269. Springer, G. F. 1967. The relation of microbes to blood-group-active substances, p. 29-47. In Cross-reacting antigens and neoantigens. Williams and Wilkins Co., Baltimore, Md. 270. Springer, G. F. 1970. Importance of bloodgroup substances in interactions between man and microbes. Ann. N.Y. Acad. Sci. 169:134-152. 271. Springer, G. F. 1971. Blood group and Forsman antigenic determinants shared between microbes and mammalian cells. Prog. Allergy 15:9-77. 272. Springer, G. F., K. K. Mittal, P. I. Terasaki, P. R. Desai, F. C. McIntire, and A. A. Hirata. 1973. Specific inhibition of HL-A alloantisera by various carbohydrate-containing macromolecules. Z. Immunitaetsforsch. Exp. Klin. Immunol. 145:166-175. 273. Stamp, L., and D. M. Stone. 1944. An agglutinogen common to certain strains of lactose and non-lactose fermenting coliform bacilli. J. Hyg. 43:266-272. 274. Staub, A. M. 1965. Somatic degraded polysaccharide of gram negative bacteria, p. 93-95. In R. L. Whistler (ed.), Methods in Carbohydrate Chemistry, vol 5. Academic Press Inc., London. 275. Stenzel, W. 1975. Zur Typologie und systematischen Stellung der intermediaren Escherichieen-Typen unter besonderer Berucksichtigung ihrer Schleimhautpathogenitat. Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. 243:97-146. 276. Stirm, S., I. 0rskov, and F. 0rskov. 1966. K88, an episome-determined protein antigen of Escherichia coli. Nature (London) 209:507508. 277. Stirm, S., F. Orskov, I. 0rskov, and A. BirchAndersen. 1967. Episome-carried surface antigen K88 of Escherichia coli. III. Morphology. J. Bacteriol. 93:740-748. 278. Stirm, S., F. Orskov, I., Orskov, and B. Mansa. 1967. Episome-carried surface antigen K88 ofEscherichia coli. H. Isolation and chemical analysis. J. Bacteriol. 93:731-739. 279. Stocker, B. A. D., and P. H. Makela. 1971. Genetic aspects of biosynthesis and structure of Salmonella lipopolysaccharide, p. 369-438. In G. Weinbaum, S. Kadis, and S. J. Ajl (ed.), Microbial toxins, vol. IV. Academic Press Inc., London. 282. Sutherland, I. W. 1969. Structural studies on colanic acid. The common lipopolysaccharide found in the Enterobacteriaceae by partial acid hydrolysis. Oligosaccharides from colanic acid. Biochem. J. 115:935-945. 283. Sutherland, I. W., and M. Korval. 1970. The synthesis of exopolysaccharide B4 Klebsiella aerogenes membrane preparations and the involvement of lipid intermediates. Biochem. J. 120:567-576. 284. Sweeney, E. J. 1970. Strains of Escherichia coli

710

ORSKOV ET AL.

associated with mortality in pigs. Ir. Vet. J. 24:108-113. 285. Takahashi, K., and S. Miura. 1968. 0 groups and antibiotic sensitivity of Escherichia coli isolated from diseased chickens. Jpn. J. Vet. Res. 16:65-72. 286. Tarcsay, L., B. Jann, and K. Jann. 1971. Immunochemistry of the K antigens of Escherichia coli. The K87 antigen from E. coli 08:K87(B?):H19. Eur. J. Biochem. 23:505-514. 287. Tatum, H. W., W. H. Ewing, and B. R. Davis. 1958. Escherichia coli 0 group 20 cultures possessing K antigen K61(B7) in common with 086:K61(B7) strains. Public Health Lab. 16:8-13. 288. Taylor, J. 1966. Host-parasite relations of Escherichia coli in man. J. Appl. Bacteriol. 29:1-12. 289. Taylor, J., and R. E. Charter. 1952. The isolation of serological types of Bacterium coli in two residential nurseries and their relation to infantile gastroenteritis. J. Pathol. Bacteriol. 64:715-728. 290. Taylor, J., and R. E. Charter. 1955. Escherichia coli 0128 causing gastroenteritis in infants. J. Clin. Pathol. 8:276-281. 291. Tixier, G., and P. Gonet. 1975. Mise en evidence d'une structure agglutinant les hematies de cheval en pr6sence de mannose, et sp6cifique des souches d'E. coli enterotoxiques d'origine bovine. C.R. Acad. Sci. Ser A 281: 1641-1644. 292. Troy, F. A., F. E. Frerman, and E. C. Heath. 1971. The biosynthesis of capsular polysaccharide in Aerobacter aerogenes. J. Biol. Chem. 246:118-133. 293. Troy, F. A., F. E. Frerman, and E. C. Heath. 1972. Synthesis of capsular polysaccharides of bacteria. Methods Enzymol. 28:602-624. 294. Turck, M., and R. G. Petersdorf. 1962. The epidemiology of non-enteric E. coli infections: prevalence of serologic groups. J. Clin. Invest. 41:1760-1765. 295. Ujvary, G. 1958. Ueber die atiologische Rolle der Escherichia coli-O-Gruppe bei extraenteral lokalisierten Infektione. Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. 170:394-422. 296. Vahlne, G. 1945. Serological typing of colon bacteria. Hakon Ohlssons Boktryckeri, Lund, Sweden. 297. Vallle, A., L. Renault, and A. Le Priol. 1970. Infections colibacillaires des veaux: etude serologique (groupes 0) de souches d'Escherichia coli isol6es en France. Bull. Acad. Vet. Fr. 43:383-389. 297a.Varga, J., and A. F. Farid. 1975. Vaccination experiments on prevention of E. coli diarrhoea in suckling calves. Acta Vet. Acad. Sci. Hung. 25:153-161. 298. Vasilieu, U. N., and I. Y. Zakharova. 1976. The structure of the determinant group in O-spe-

BACTERIOL REV.

cific polysaccharide of E. coli 020:K84:H34. Bioorg. Chem. 2:199-206. 299. Villemez, C. L., and A. F. Clark. 1969. A particle bound intermediate in the biosynthesis of plant cell-wall polysaccharide. Biochem. Biophys. Res. Commun. 36:57-63. 300. Voll, M. J., and L. Leive. 1970. Release of lipopolysaccharide in Escherichia coli resistant to the permeability increase induced by ethylene diamine-tetraacetate. J. Biol. Chem. 245:1108-1114. 301. Wagner, N. 1971. Untersuchungen uber die Haufigkeitsverteilung von E. coli-serogruppen aus der Faecalflora gesunder Erwachsener zweier geographisch getrennter Untersuchungsgruppen. Thesis, University of Frankfurt, Frankfurt, West Germany. 302. Wallenfels, B., and K. Jann. 1974. The action of bacteriophage fi 8 in two strains ofEscherichia coli 08. J. Gen. Microbiol. 81:131-144. 303. Westphal, O., and K. Jann. 1965. Bacterial lipopolysaccharides. Extraction with phenolwater and further applications of the procedure, p. 83-91. In R. L. Whistler (ed.), Methods in Carbohydrate Chemistry, vol. 5. Academic Press Inc., London. 304. Westphal, O., 0. Luderitz, and F. Bister. 1952. Ueber die Extraktion von Bakterien mit Phenol/Wasser. Z. Naturforsch. Teil B 7:148-155. 305. Wiedemann, B., and H. Knothe. 1969. Untersuchungen uber die Stabilitat der Koliflora des gesunden Menschen. Arch. Hyg. 4:342352. 306. Wiley, B. B., and H. W. Scherp. 1958. Capsular polysaccharides of Escherichia coli types K28(A) and K34(A). Can. J. Microbiol. 4:505-566. 307. Winkle, S., M. Refai, and R. Rohde. 1972. On the antigenic relationship of "Vibrio cholerae" to "Enterobacteriaceae". Ann. Inst. Pasteur Paris 123:775-781. 308. Wittig, W. 1965. Zum Vorkommen des K-Antigens 88(L) bei Escherichia coli-Stammen von Schweinen. Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. 197:487-499. 309. Wramby, G. 1948. Investigations into the antigenic structure of Bact. coli isolated from colon. With special reference to coli septicaemia. Thesis, Appelbergs Boktryckeri AB, Uppsala, Sweden. 310. Wright, A., M. Dankert, P. Tennessey, and P. W. Robbins. 1967. Characterization of a polyisoprenoid compound function Ac in O-antigen synthesis. Proc. Natl. Acad. Sci. U.S.A. 57:1798-1803. 311. Zamenhof, S., G. Leidy, P. L. Fitzgerald, H. E. Alexander, and E. Chargaff. 1953. Polyribophosphate, the type-specific substance of Hemophilus influenzae, type b. J. Biochem. 203:695-704.

Protective capacity of antibodies against Escherichia coli and K antigens.

Genetic transfer of Salmonella O antigens to Escherichia coli O8.

Local immunie response in experimental pyelonephritis in the rabbit. III. Lymphocyte responsivieness to O and K antigens of Escherichia coli.

Behavior of Escherichia coli K antigens K88ab, K88ac, and K88ad in immunoelectrophoresis, double diffusion, and hemagglutination.

Pathogenicity and molecular genetics of O-specific side-chain lipopolysaccharides of Escherichia coli.

The antigenic relationship between Candida tropicalis and Escherichia coli (O 75, O 163) and Salmonella serotype Aberdeen O antigens.

Antibody response to Escherichia coli O antigens and influence of the protein moiety of the endotoxin.

Role of Escherichia coli K capsular antigens during complement activation, C3 fixation, and opsonization.

Serologic responses to somatic O and colonization-factor antigens of enterotoxigenic Escherichia coli in travelers.

The major histocompatibility complex of rhesus monkeys. VI. Serology and genetics of Ia-like antigens.

Interactions of outer membrane proteins O-8 and O-9 with peptidoglycan sacculus of Escherichia coli K-12.

The local immune response to Escherichia coli O and K antigen in experimental pyelonephritis.

Genetics of nitrofurazone resistance in Escherichia coli.

Screening for cross-reacting capsular polysaccharide K antigens of Escherichia coli using antiserum agar.

Complement activation in relation to protective function of antibodies to Escherichia coli O antigens.

Relative opsonic and protective activities of antibodies against K1, O and lipid A antigens of Escherichia coli.

Genetics and enzymology of DNA replication in Escherichia coli.

Evaluation of methods for the determination of Q and K antigens of an 02:K1(L) strain of Escherichia coli.

Influence of molecular size and osmolarity of sugars and dextrans on the synthesis of outer membrane proteins O-8 and O-9 of Escherichia coli K-12.

Mapping of nrdA and nrdB in Escherichia coli K-12.

The cell-wall lipopolysaccharide of Escherichia coli K-12. Structure and acceptor site for O-antigen and other substituents.

Development of the ECODAB into a relational database for Escherichia coli O-antigens and other bacterial polysaccharides.

Novel antigens for enterotoxigenic Escherichia coli vaccines.

The major histocompatibility complex of Rhesus monkeys IX. Current concepts of serology and genetics of Ia antigens.